PHOTOMOD. Block Adjustment PDF

Digital Photogrammetric System
Version 5.21
USER
MANUAL
Block Adjustment
1.
PHOTOMOD Solver. Module of the block adjustment ....................................................... 5
2.
Installation and starting ........................................................................................................ 5
3.
Adjustment of projects of different types ........................................................................... 5
4.
Coordinate systems .............................................................................................................. 6
5.
Free model .............................................................................................................................. 7
6.
Source window ...................................................................................................................... 7

6.1
Points displaying ...................................................................................................................... 8
6.2
View parameters ...................................................................................................................... 9
6.3
Points selection by filter ......................................................................................................... 10
6.4
Points window ........................................................................................................................ 12
6.5
Images window ...................................................................................................................... 13
6.6
Stereopairs window ............................................................................................................... 13
6.7
Information window on the Source tab .................................................................................. 13
6.8
Images, stereopairs and strips displaying ............................................................................. 15
6.9
Cursor modes ........................................................................................................................ 15
6.10
Graphic windows toolbar ....................................................................................................... 16
7.
Methods of block adjustment ............................................................................................. 17
8.
Parameters window ............................................................................................................. 18

8.1
Coordinate system tab ........................................................................................................... 18
8.2
Adjustment tab ....................................................................................................................... 18
8.2.1 Independent strips method .................................................................................................... 20
8.2.2 Independent stereopairs method ........................................................................................... 21
8.2.3 Bundle adjustment ................................................................................................................. 21
8.2.4 Systematic error compensation ............................................................................................. 23
8.2.5 Select subblock ...................................................................................................................... 24
8.3
Report tab .............................................................................................................................. 25
9.
Computations ....................................................................................................................... 36
10.
Stereopairs tab ..................................................................................................................... 38
10.1
Error vectors displaying ......................................................................................................... 38
10.2
Residuals window .................................................................................................................. 39
Block adjustment
2011
10.3
Brief residuals report on Stereopairs tab ............................................................................... 40
10.4
Tie residuals window ............................................................................................................. 41
10.5
Information window on the Stereopairs tab ........................................................................... 43
10.6
Editing points ......................................................................................................................... 45
11.
Images tab ............................................................................................................................ 47
11.1
Residuals window .................................................................................................................. 48
11.2
Residuals on image ............................................................................................................... 49
11.3
Brief residuals report on the Images tab................................................................................ 50
11.4
Information window on the Images tab .................................................................................. 50
12.
Report creation .................................................................................................................... 51
13.
Console window .................................................................................................................. 53
14.
Saving the block scheme .................................................................................................... 53
15.
Finishing the work in PHOTOMOD Solver ........................................................................ 53
16.
Block adjustment by external programs ........................................................................... 53
17.
Single aerial image adjustment .......................................................................................... 56
18.
Adjustment of scanner blocks ........................................................................................... 56
Appendix A. Exterior orientation angular systems ......................................................................... 64

Appendix B. Recommendations on aerial images blocks adjustment.......................................... 66
1.
Typical errors ....................................................................................................................... 66
2.
Stages of control ................................................................................................................. 66

2.1
Interior orientation control ...................................................................................................... 66
2.1.1 Interior orientation control ...................................................................................................... 66
2.2
Relative orientation control .................................................................................................... 66
2.2.1 Points recognition control ...................................................................................................... 66
2.2.2 Control of uniform distribution of points measurements on stereopairs ................................ 67
2.2.3 Control of vertical parallaxes on stereopairs ......................................................................... 67
2.3
Control of triplets .................................................................................................................... 67
2.3.1 Control of uniform distribution of points measurements in triplets ........................................ 67
2.3.2 Control in triplets .................................................................................................................... 68
2.3.3 One strip adjustment in free model using independent models method ............................... 68
2.4
Control of free model adjustment .......................................................................................... 68
2.4.1 Using of splitting on sub-blocks ............................................................................................. 68
RACURS Co., Ul. Yaroslavskaya, 13-A, office 15, 129366, Moscow, Russia
PHOTOMOD 5.21
2.4.2 Detection of blunders in camera focus input ......................................................................... 68

2.4.3 Detection of blunders in strips tying....................................................................................... 68
2.4.4 Intermediate control of tie measurements blunders .............................................................. 69
2.4.5 Blunders detection on interstrip not-transferred points ......................................................... 69
2.4.6 How to detect errors of tie points measurements .................................................................. 69
2.4.7 How to detect errors of automatic tie points measurements ................................................. 70
2.5
Control of adjustment of georeferenced blocks ..................................................................... 70
2.5.1 How to detect a reference to wrong coordinate system ........................................................ 70
2.5.2 Detection of blunders in GCP measurements ....................................................................... 71
2.5.3 Selection from several adjacent GCP, if their error' vectors have opposite direction ........... 71
2.5.4 Control of GCP located on one stereopair ............................................................................ 71
2.5.5 Control of adjustment with projection centers ....................................................................... 71
2.5.6 Detection of systematic errors of interior orientation ............................................................. 72
2.5.7 Bundle adjustment and adjustment import: errors on points measured on one stereopair .. 73
2.6
Control during points filtering in PHOTOMOD Solver ........................................................... 73
2.6.1 Irregularity of points' distribution control ................................................................................ 73
3.
Bundle adjustment specific features ................................................................................. 73
4.
Interpretation of errors shown in PHOTOMOD Solver on Stereopairs and Images tabs

74
5.
Processing of survey with big images overlap ................................................................ 74
6.
Recommendations on long-focus camera survey ........................................................... 75
7.
Tie points measuring in forested areas, in areas with bad stereo effect. Tie points
measuring between strips acquired in different seasons .............................................................. 75
8.
2011
Tie points measuring on lengthy objects boundaries ..................................................... 76
Block adjustment
2011
1. PHOTOMOD Solver. Module of the block adjustment

PHOTOMOD Solver is designed for adjustment of blocks of aerial and satellite images.
Interior and relative orientation of block images must have been carried before adjustment
(see chapter Aerial triangulation).
Besides, you can export and import data from/to PHOTOMOD Solver module to/from
external software of data adjustment, see the chapter Block adjustment by external
programs.
2. Installation and starting

PHOTOMOD Solver is started with menu command Adjustment | Block adjustment or
pressing the button
on PHOTOMOD Core main toolbar.
If necessary, user can start the adjustment process for the selected part of the block images.
Selecting sub-block is performed using PHOTOMOD Solver means, or by menu command
Adjustment | Select subblock in PHOTOMOD Core.
3. Adjustment of projects of different types

PHOTOMOD system works with two types of projects (based on the source image geometry)
Central projection (images obtained by photo cameras) and Satellite scanner imagery
(images obtained by scanning devices such as SPOT, ASTER, IRS, IKONOS etc.). The
project type is selected in PHOTOMOD Core module during project creating (see the user
manual Project creation).
The process of adjustment for blocks of airborne images is performed as described in
chapters 6-18 of present User Manual.
PHOTOMOD 5.21
Project type selection

Scanner blocks adjustment is described in the corresponding chapter.
Images acquired by airborne digital sensor ADS 40 (ADS 80) by Leica Geosystems,
Swizerland, after adjustment with the ORIMA software, can be processed in PHOTOMOD
system.
4. Coordinate systems
The coordinate system can be selected either in PHOTOMOD Core shell during the project
creation, or in PHOTOMOD Solver during block adjustment.
In some cases images orientation and their further processing in PHOTOMOD system are
executing without images referencing to ground control points (without referencing to
coordinate system, that corresponds the free model term, see the chapter Free model).
2011
Block adjustment
2011
5. Free model
Creating of the free model is used in the absence of GCPs at the time of block adjustment. In
this case, only interior orientation and measurement of tie points in stereo pairs and between
strips is performed. PHOTOMOD Solver builds a model, counting only tie point residuals.
This make possible building terrain models and creating 3D vector objects in PHOTOMOD
Core. Thus, by the time of getting GCP's coordinates the majority of work with project may be
complete. When GCPs are available, it is necessary to measure the GCPs and readjust the
block in the selected coordinate system in the PHOTOMOD Solver, resulting in all vectors
and TIN's created in the free model being transformed into the selected geodetic coordinate
system.
To adjust block in free model select the option Free model in Parameters | Adjustment
window and set the approximate aerial survey basis in units of coordinate system of a project
(meters, feet, etc., in Basis field). Free model errors are displayed in the same units and
depend on the basis accuracy.
Note. There is no way to create orthomosaic and contour lines in free model
6. Source window
of
Once PHOTOMOD Solver is started for the project at the first time (using the button
the PHOTOMOD Core program), opens the Block adjustment main window on the Source
tab, and performed automatic loading of source data from the interior orientation and
measurement of tie points in stereo pairs and between strips.
When adjustment is completed you get three tabs in the graphic window: Source,
Stereopairs and Images. First of them shows the initial images block, the second and the
third ones the block adjusted with residuals vectors, that makes easy visual analysis of
adjustment results. After adjustment procedure the block scheme is shown on the
Stereopairs tab of the main window (see the chapter Stereopairs tab), and also on the third
tab Images, see the chapter Images tab.
In the upper part of PHOTOMOD Solver window there are main buttons that allow to perform
the adjustment procedure:
opens the window for adjustment parameters setting (see the chapter
Parameters window)
starts the process of computing of exterior orientation parameters and

coordinates of all points of a block (see the chapter Computations)
opens the window with adjustment results (see the chapter Report
creation)
saves the adjustment results that allows to proceed to next project
processing stage in PHOTOMOD system (see the chapter Finishing the work in
PHOTOMOD Solver)
exits PHOTOMOD Solver
PHOTOMOD 5.21
Initial block state on the Source tab

If some errors occur while image block loading you will receive an appropriate message and
further work will be blocked.
If the source dataset is loaded successfully, the Parameters button becomes available. You
have to set process parameters if you start Block adjustment for the first time (see the
chapter Parameters window). Parameters are memorized when exiting PHOTOMOD Solver
and restored for each future block adjustment session.
Borders of selected images, stereopairs and strips are shown in lighter color.
6.1
Points displaying
Points are shown on block scheme by graphic symbols. Point type is shown by the color and
shape of symbol as follows:
points with known coordinates (control and check):
-
XYZ points double triangle

XY points single triangle;
Z points circle;
ground control red color
check black color;
rejected from adjustment white color
projection centers small black square

tie points
2011
;
;
measured on at least one stereopair small diagonal cross

8
Block adjustment
2011
measured on no stereopairs small circle
measured on only one stereopair black color
measured on more than one stereopair on one strip green color (for example,
measured in triplet)
interstrip point deep blue color
rejected from adjustment white color

measured on only one image on one strip magenta color (light magenta if rejected)
ground point (tie point that is not targeted and is not a projection center) no symbol;
targeted point bold symbol (one of described above)
;
-
(light blue if rejected);
Any point can be displayed using a set of symbols. For example, control XYZ point, identified
on two strips (in the overlap areas) is displayed as double red triangle and blue diagonal
cross
. All projection centers except the first and last ones are displayed as a square and
a diagonal cross
The white circle shows selected points.

Symbols for points displaying are also described in the chapter Points selection by filter.
6.2
View parameters
The window View parameters (opened by the icon

) is used to select the elements to be
included to the block scheme. The window is divided into five main panes, related to
displaying points, images, stereopairs, strips and the map grid respectively.
View parameters
9
PHOTOMOD 5.21
Points panel
- Show / all all points measured on the images of the block are displayed.
- Show / filter points to be displayed are selected according to filter settings. Click the
Parameters button (if the Filter option is selected) to open a window used for setup
points filter, see the chapter Points selection by filter. Thus, points selected in this
window are visualized in viewing window.
- Names defines which points will be displayed with labels (options All, Selected,
None).
Point symbol width - allows changing the thickness of symbols representing interstrip
points
Images panel
- Show shows / hides image borders.
- Names defines which images will be displayed with labels (options All, Selected,
None).
Stereopairs panel
- Show shows / hides borders of stereopairs.
- Names defines which stereopairs will be displayed with labels (options All, Selected,
None).
Strips panel
- Show shows / hides borders of strips.
- Names defines which strips will be displayed with labels (options All, Selected,
None).
Grid panel allows to show map grid lines (lines) and coordinates (coordinates).
6.3
Points selection by filter
) is used to filter and select points

The dialog Selection by filter (opened by the icon
based on some criteria and to show selected points using symbols on block scheme. The
symbols are described in the chapter Points displaying.
2011
10
Block adjustment
2011
Select points window

There are five panels, used to setup different criteria:
By coordinates. List of point types used to select/deselect points includes:
ground control ground control points, projection centers
check check points, projection centers
rejected ground points excluded from adjustment, with known coordinates (ground
control, check and projection centers, both in plane and height (3D), plane
(horizontal) and height (vertical)),
no coordinates points with unknown coordinates (tie, targeted and projection
centers (if their coordinates are not known))
By ties
single tie points measured on no stereopairs
stereopair tie points measured at least on one stereopair
triplet tie points measured at least on one triplet
strip tie tie points measured in strips overlap
non-transferred tie points not measured on the adjacent images
used tie point used for the block adjustment as ties
rejected tie point rejected from the block adjustment as ties
Auxiliary
geodetic tie points that are neither targeted nor projection centers
targeted targeted points
projection centers projection centers
with changed type points which type was changed in PHOTOMOD Solver
By residual value
worst points selection of points with maximum residual values. Number of worst
points is to be set in corresponding text field.
11
PHOTOMOD 5.21
over acceptable residuals selection of points with residual values greater than the
Acceptable residual parameter, indicated in the Parameters window, on the
Report tab, in Acceptable residuals panel, see the chapter Report tab.
You can set on ground and/or tie options to select ground and/or tie points that satisfy
chosen criteria as well as XY and Z options to take into account XY and/or Z residuals only.
In selected models
all select points, which belong to all models
at least one select points, which belong to at least one model
Also invisible mark the option to select both visible and invisible points. If the option is
not marked, only visible points will be selected. Points may be invisible, if they do not
satisfy to preset filter of points displaying.
Note. The panel By residual value is inactive if the window is opened from the Source
tab but is used on the Stereopairs and Images tabs
The buttons Select, Add, Toggle and Subtract are used to execute the appropriate
selection operations with selected points.
6.4
Points window
Points window (opened by the icon

the block scheme by their names.
point list) is used for quick selection of points in
Points window
2011
12
Block adjustment
2011
At the upper part of the window there is a text field, where you can type the point name to be
searched and selected.
You can also select point in the point list below (sorted in alphabetical order). If the Auto
option is on, selected point is automatically marked in the block scheme (outlined by the
white circle
). If you need to select several points, turn off the Auto option and use the
Select button for points selection. If current point is already selected, click the Select button
to cancel point selection. The option List of selected objects is used to show a list of
currently selected objects.
6.5
Images window
image list) is used for quick selection of images in

Images window (opened by the icon
the block scheme by their names. At the upper part of window there is a text field, where you
can type the image name to be searched and selected. You can also select image in the
image list below (sorted in alphabetical order).
If the Auto option is on, selected image is automatically marked in the block scheme
(bordered by yellow bold line). If you need to select several images turn off the Auto option
and use the Select button for images selection. If current image is already selected, click the
Select button to cancel image selection. The option List of selected objects is used to
show a list of currently selected objects.
Note: in order for "Images" window to display the list of images, it is necessary to
check the option "show" in the "Images" pane of the "View parameters" window.
6.6
Stereopairs window
stereopair list) is used for quick selection of

Stereopairs window (opened by the icon
stereopairs in the block scheme by their names. At the upper part of window there is a text
field, where you can type the stereopair name to be searched and selected. You can also
select stereopair in the list below (sorted in alphabetical order).
If the Auto option is on, selected stereopair is automatically marked in the block scheme
(bordered by dark-blue bold line). If you need to select several stereopairs, turn off the Auto
option and use the Select button for stereopairs selection. If current stereopair is already
selected, click the Select button to cancel stereopair selection. The option List of selected
objects is used to show a list of currently selected objects.
Note: in order for "Stereo pairs" window to display the list of stereo pairs, it is
necessary to check the option "show" in the "Stereo pairs" pane of the "View
parameters" window.
6.7
Information window on the Source tab
point attributes) is used to view the attributes

Information window (opened by the icon
of currently selected point: its type, coordinates and measurements. There is a text field at
the upper window part, which shows the name of selected point, if the only one point is
selected.
13
PHOTOMOD 5.21
Mark the option Editing panel to open the panel allowing to change point type (control or
check and vice versa), and also to exclude points from the adjustment.
Current points type is indicated by marked check box near the appropriate type name in this
panel.
To change the type of the selected point for the next adjustment, push one of the appropriate
buttons:
- control
- check
- rejected as a ground control or check point, but can be used as a tie
You can edit control point coordinates using the window opened by the button
(Edit coordinates), which is used to open the window
for editing coordinates of triangulation points, described in the chapter Editing points.
Information window shows the data for selected point

Panel By ties contains the button
(Edit tie points),
used for tie points editing in special window. See the chapter Editing points.
You can exclude tie point selected from the further adjustment by the button
excluded tie point to adjustment push the button
restoring push the buttons
2011
and
. To restore
. For ground control and check points
accordingly).
14
Block adjustment
2011
The rest part of the Information window is used to display selected point attributes. It
contains information if the only one point is selected in graphic window.
Geodetic. Shows initial (real) coordinates entered in PHOTOMOD AT module

(reference coordinates) and also initial weights of coordinates (reference weights).
Ties. Used to show information about tie points. Images number of images, where
the point is measured, and measurements of the point (its coordinates) in these
images, number of stereopairs and strips.
Type indicates whether the point was ground control, check or rejected during
adjustment, and whether it used as a tie. It also indicates is the point a projection
center, and shows point code (or image code for projection centers)
Images list of images the point was measured on
Stereopairs list of stereopairs the point was measured on
Strips list of strips the point was measured on (on one image at least).
The window allows to select images, stereopairs and strips where the point is measured, in
the following way.
Holding Ctrl key pressed you can click the following lines in the window: Tieimages,
Tieimagesimage
name,
Images,
Imagesimage
name,
Stereopair,
Stereopairsstereopair name, Strips, Stripsstrip name. At that the appropriate object or
object group will be highlighted. If the object is already selected, the selection will be
cancelled. Mouse selection while holding Ctrl and Shift keys inverts selection.
6.8
Images, stereopairs and strips displaying
Images, stereopairs and strips are shown on the block scheme in the following way:
Images are shown by yellow frames, selected images by light yellow frames
Stereopairs (images overlap areas) by blue frames
Strips (the whole areas of coverage of strip's images) by magenta frames.
Models and images names can also be displayed if the names option in View parameters
window is on (see the chapter View parameters).
6.9
Cursor modes
select mode is
Selection mode is the default cursor mode (is on when the button
pushed). You can define the type of objects to be selected by following icons: Select points
, Select images
and Select models
. Click the object neighborhood to select a
closest object (the screen distance between the cursor and the object should be less than
the cursor size).
Note. Points are selected by their centers, images, stereopairs and strips by their
frames
By default, selection of objects cancels the selection of previously selected ones (if they were
selected in Points selection, Images selection and Model selection modes). All other
objects (if any) remain selected, since there is a possibility of using objects filter that depends
on selected objects (see the chapter Points window). To add objects to selection, select them
by mouse along with pressed down Ctrl key. In this case the cursor gets plus sign added
. To deselect some objects press and hold Shift key. In this case the cursor gets
15
PHOTOMOD 5.21
minus sign added
. To invert selection of the group of objects press and hold down
Shift and Ctrl keys. At that, the cursor gets minus-plus sign added
(Unselect objects) to unselect all
part of the window or push upper menu button

objects.
Pan mode is activated by pushing the Pan icon
. Click the blank
or by pressing and holding down the
Space bar if you are in Select mode (when the button
is pushed). Once the pan mode
is on, small pictogram with arrows appears nearby the cursor

. There are following
options in the pan mode:
- panning the image (the default one). Press and hold mouse button to pan over the
image.
- zoom in. Press Ctrl key to zoom in the image using the zoom box (drawn by the
mouse). Mouse click causes two times image enlarging. Or press and hold
-
Ctrl-Space shortcut if you are in Selection mode (when the button

is pushed).
zoom out. Press Shift or Alt keys to zoom out the image in the same way as
described above. Or press and hold Shift-Space or Alt-Space if you are in Selection
mode (when the button
is pushed).
6.10 Graphic windows toolbar

Graphic windows toolbar contains the following buttons:
Source scale to go back to original block scheme zoom
Zoom in (duplicated by * key) turns on the mode which allows to enlarge the
block scheme twice by each mouse click
Zoom Out (duplicated by / key) turns on the mode which allows to reduce the
block scheme twice by each mouse click
Pan mode allows to pan the block scheme by mouse. See the chapter Cursor
modes
Select mode turns on selection mode for points, images, stereopairs and strips.
See the chapter Cursor modes
Select points allows to select points. See the chapter Points displaying
Select images allows to select images. See the chapter Images, stereopairs and
strips displaying
Select models allows to select stereopairs and strips. See the chapter Images,
stereopairs and strips displaying
Selection by filter opens/closes a dialog used to filter and select points (see the
chapter Points selection by filter)
Unselect objects cancels all selection
Marker parameters opens dialog used to change cursors shape, size and color.
2011
16
Block adjustment
2011
View parameters opens a dialog used to set view parameters (see the chapter
View parameters)
Residuals opens a dialog used to set parameters of residuals viewing. See the
chapter Residuals window, Residuals window. Available only on the Stereopairs and
Images tabs.
Point list opens/closes a window with a point list (see the chapter Points window)
Point attributes opens/closes a window with the information, related to selected

point. See the chapter Information window on the Source tab, Information window on the
Stereopairs tab, Information window on the Images tab.
Points residuals opens / closes a window for displaying error vectors on selected
tie point (see the chapter Tie residuals window). Available only on the Models tab.
- Image residuals opens / closes a window for displaying error vectors on selected
image (see the chapter Residuals on image). Available only on the Images tab
Image list opens/closes Images window with an image list (see the chapter
Images window)
Stereopair list opens/closes Stereopairs window with a model list (see the
chapter Stereopairs window)
Residuals window opens/closes a window containing RMS, mean absolute and

maximum adjustment error values. Used for quick estimation of the adjustment results
without opening a full report. Differs for Stereopairs and Images tabs (and is available
only on these tabs). See the chapter Brief residuals report on the Stereopairs tab and
Brief residuals report on the Images tab
Hide windows closes all additional windows

Save block scheme to the vector (wmf) or raster (bmp) format. See the chapter
Saving the block scheme.
7. Methods of block adjustment

PHOTOMOD Solver provides three algorithms to be used for the block adjustment
procedure: Independent strips, Independent stereo pairs and Bundle adjustment.
Independent strips. This method is used to check out the gross errors, such as wrong
coordinate values of control points, incorrect tie point's measurements, etc. Accuracy of
this method in case of long strips (more than 10 images) may be dozen times worse than
adjustment accuracy achieved by the other two methods. See method setup parameters
in the chapter Independent strips method
Independent stereo pairs. Use this method to increase the accuracy, you have
achieved by Independent strips method, and to detect more delicate errors and possibly
for final adjustment. See method setup parameters in the chapter Independent
stereopairs
Bundle adjustment is used for the final block adjustment. See details about method
setup in the chapter Bundle adjustment.
In most cases, it is practical to perform adjustment alternately by method of independent

stereopairs and bundle adjustment for searching and correction of delicate errors.
17
PHOTOMOD 5.21
You can select adjustment method and setup it on the Adjustment tab of the Parameters
window opened by pushing the button
, see the chapter Adjustment tab.
8. Parameters window
Parameters window opens after clicking the button
report parameters) and includes the following tabs:
8.1
(Edit adjustment and
Coordinate system tab
The Coordinate system tab is used to select the project coordinate system (see the chapter
Coordinate systems for details). Push the Change button and select coordinate system and
projection in standard PHOTOMOD window, see Project creation User Manual. After
choosing or changing the coordinate system and map projection for images block processing
you can adjust the block.
8.2
Adjustment tab
The Adjustment tab is used to select the Block adjustment method: Independent strips,
Independent stereopairs or Bundle adjustment (see the chapter Methods of block
adjustment).
Use Free model adjustment mode to adjust a block without GCP, and to provide following
georeference. The Free model option is used this procedure (see the chapter Free model).
2011
18
Block adjustment
2011
Adjustment parameters
Free model adjustment is executed using distance between projection centers, indicated in
Basis field (in units of coordinate system of a project). Basis value by default is 1000.
You can reject the worst tie points during adjustment automatically, using filtering options on
the Tie point rejection panel: worst points (number of points with maximal residuals) and
over acceptable residuals (all points with residuals exceeded the accepted values indicated
in Acceptable residuals panel on Report tab, see the chapter Report tab). If the options are
on, filtered points will be excluded from adjustment.
When the adjustment method is selected you can setup it in the window, opened by the
button Method parameters, see the chapter Independent strips method, Independent
stereopairs method, Bundle adjustment.
Systematic errors of adjustment are considered, and you can setup their compensation
parameters in the window opened by the button Systematic error compensation, see the
chapter Systematic error compensation.
It is possible to select subblock without exiting PHOTOMOD Solver. In order to do it, it is
necessary to press the Select subblock button, see chapter Selecting subblock. For the
adjustment of selected subblock, it is necessary to check the Adjust subblock option
19
PHOTOMOD 5.21
8.2.1
Independent strips method
When you have selected independent strips method of adjustment, you can setup the
process parameters in the window opened by the button Method parameters on the
Adjustment tab (see the chapter Adjustment tab).
Independent strips method setup

The following parameters are available: Equation weights for ground control, tie points
and projection centers with specified coordinates. You can reduce the residuals values by
changing the equation weights. However, real adjustment accuracy can be also reduced in
this case. In many cases the weight value of 1 is near to optimal. You have to have check
points enough to control the result of optimal value search. If there is a lack of check points
you can temporarily transfer some GCP to check. The decimal delimiter is "." (period). The
exponent sign is "e". Example: "5.7e-4" = 0.00057.
The slider Adjustment accuracy is used to setup relative accuracy, and when it is reached
the iteration process of adjustment may be completed.
You are recommended to use the default value 0.5. When it is 0.5-0.55, the adjustment
process converges and provides an optimal ratio accuracy-time. If you increase the value,
the accuracy generally does not improve considerably, but processing time may increase
very much up to infinite looping of iterations (in this case you may stop iterations when you
need using the Stop button). If the accuracy value is less then 0.5 the adjustment time is
reducing due to accuracy decreasing. The algorithm of optimization has a complex nature,
that is why the dependence of adjustment time and accuracy from this parameter value is
nonlinear.
When processing close range photogrammetry projects, containing points with big
differences in x-parallaxes, the option Maximum distance (in bases) is used. Points located
from projection centers on the distance more then indicated in this field are excluded from
adjustment on each stereopair automatically. The distance is input in survey bases, i.e. if
stereopair basis is 10 m and in this field the value 100 is set up, then the points located
farther then 1000 m from projection centers will be excluded.
It is recommended to set up the values not more than 100, because when the points with
distance more than 100 survey bases are used for adjustment, the method of independent
strips (or stereopairs) would fail or return a blunder result.
Basically (if there are no blunders in measurements), reduction of this value leads to
decreasing of RMS for the whole block. However if too small value is set up, there will be
2011
20
Block adjustment
2011
very few tie points (or no points at all) in triple overlap, that will result in adjustment error with
an appropriate message.
During adjustment with method of independent strips, systematic errors are accumulated
along the strips, which leads to large adjustment residuals when the length of a strip is more
than 20-30 images. In the Maximum strip length it is recommended to set the value of 1030 images. Longer strips are broken into parts of the said size during adjustment.
8.2.2
Independent stereopairs method
Parameters of independent stereopairs method are setup similarly to those of independent

strips method (see the chapter Independent strips method). However, there is a possibility to
choose the algorithm of adjustment optimization in the following way.
Independent stereopairs method setup

In settings window there is the Optimization algorithm panel, that allows to select one of
the following methods of adjustment optimization: Gradient (by default) and Matrix
inversion. The last method was used in previous (up to 4.1 inclusive) versions of
PHOTOMOD system. It was developed for moderate blocks of images (less than 500
images).
8.2.3
Bundle adjustment
When using bundle adjustment method you can setup its parameters in the window opened
by pushing the button Method parameters on the Adjustment tab, see the chapter 8.2
Adjustment tab.
21
PHOTOMOD 5.21
Bundle adjustment setup

Panel Measurement precision (RMS) is used to configure the following parameters:
- Image measurements - RMS accuracy of the identification of points on images (see user
manual Block measurement). It is measured in mm or pixels.
This accuracy depends on the quality of photomaterial (photographic grain, blur, image
deformations that were not considered in camera parameters, including geometric distortions
of scanner) and on accuracy of recognition performed automatically or manually. In case of
manual measurements the main factors include stereovisual acuity, terrain type, image
zoom, carefulness of work. For automatic recognition settings and terrain type are important.
If scanning resolution of image corresponds to image's quality, and points are recognized
properly, this value has to be between 0.51.0 of pixel size.
Weights panel is used for setting the following parameters:
- ground control points - weight of GCP's coordinates. Setting a value W for weights is
equivalent to multiplying the numerical value of the accuracy of GCP in the catalog (see user
manual Block measurement) by 1 / W. Setting the value W> 1 strengthens binding of
triangulation network to GCP's coordinates, W <1 - weakens.
- projection centers - the same as weight for GCP's, but for the coordinates of the
projection centers. Accuracies of coordinates of projection centers are specified in the
catalog of exterior orientation elements (see User manual Block measurement).
- exterior orientation angles (omega, phi, kappa) - the same as weight for GCP's, but for
angles of exterior orientation. Accuracy of exterior orientation angles are set in the catalog
of exterior orientation elements (see user manual Block measurement).
The option Use exterior orientation angles enables using exterior orientation angles in
adjustment procedure.
2011
22
Block adjustment
2011
The Adjustment accuracy slider sets the relative accuracy, reaching which stops the
iterative adjustment process. The same as Adjustment accuracy for method of independent
strips (see Independent strips method).
8.2.4
Systematic error compensation
For automatic calculation and compensation of some systematic errors of block adjustment
you can use settings in a window opened by the button Systematic error compensation on
the Adjustment tab, see the chapter Adjustment tab.
In the Projection center coordinates (GPS) panel you can select a type of compensation
applied to projection center coordinates (GPS).
If the airborne survey was performed using GPS equipment, for measuring the projection
centers coordinates, you can calculate the corrections to these coordinates during
adjustment, by choosing the appropriate option in the Compensation type panel. Usually
the corrections of GPS device, which were introduced on strip survey should be considered.
So it is recommended to turn on the option common for strip. Calculated corrections of
coordinates measured by GPS could be included into adjustment report, see the chapter 8.3
Report tab. Coordinates corrections for projection centers are calculated by ground control
points and are as follows:
X = G 0x + G 1x N
Y = G 0y + G 1y N ,
Z = G 0z + G 1z N
23
PHOTOMOD 5.21
where N ordinal number of image in block or strip, or time corresponding to the GPSmeasurement of the image center (defined by the Dependency (for the 1st order
polynomial) option), G polynomial coefficients that are calculated during adjustment. If
polynomial order equals 1 (input in the Polynomial degree panel), then both G 0 , and G 1 are
used, if the order is 0, then just G 0 are used, and G 1 are assumed to be equal 0.
The panel Exterior orientation angles (constant rotation) is used to setup the type of
corrections that are applied to exterior orientation angles of images, if they are in the project
and are used for adjustment. These parameters are available only for bundle adjustment
method and are as follows:
where A coefficients calculated during adjustment.

The type of polynomial corrections for adjustment results is setup in the Polynomial
corrections for block panel: XY, Z, XYZ (at user's option). The panel is available only for
methods of independent strips and independent stereopairs.
Polynomial corrections are calculated using the following formulas:
X = Ax + B x X + C x Y + D x X 2 + E x XY + Fx Y 2
Y = Ay + B y X + C y Y + D y X 2 + E y XY + Fy Y 2 .
Z = Az + B z X + C z Y + D z X 2 + E z XY + Fz Y 2
A , , F coefficients are calculated during post-processing of adjustment results

performed by GCP (including projection centers). Corrections obtained are added to adjusted
coordinates of all points of the block, including projection centers.
In most cases you are recommended to turn off polynomial corrections usage.
8.2.5
Select subblock
It is possible to select subblock without exiting PHOTOMOD Solver. In order to do it, it is

necessary to press the Select subblock button.
2011
24
Block adjustment
2011
There are various ways to select images. Separate images can be selected by checking
checkboxes on the left. Strips may be selected at once by checking checkboxes for the
strips. Also the following tools can be used:
- Select all the images in the list;
- Deselect all images in the list;
- Invert the selection of images;
- Select the highlighted images;
- Deselect highlighted images;
- Select images which are selected in the block scheme in 2D window;
- Select in a blocks scheme in 2D window the images which are selected in the list.
8.3
Report tab
The Report tab is used to select adjustment report parameters and parameters of residuals
visualization on the block scheme (acceptable residuals and method of tie residuals
calculating).
25
PHOTOMOD 5.21
Report parameters
The Include in the report panel allows you to include the following values to the report.
Residuals the option allows to include the section with adjustment residuals into the report:
----------------------------------------------------------------------------------------------------------------------------------------------------------------------Estimation of block adjustment accuracy
Unacceptable residuals are marked with "*".
----------------------------------------------------------------------------------------------------------------------------------------------------------------------Generalized adjustment residuals information
Sigma_0 = 0.623
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------
Sigma_0 value is calculated only for bundle adjustment method. It is calculated using the
following formula:
Nt im 2 im 2 N g
X ig
1
xi y i
=
+
+
N t + N g i=1 s x,imi s im
i=1 s X,i
y,i

g
im
im
where x , y residuals on images, X ,
2
0
Yi g
+ g
s
Y,i
Z ig
+ g
s
Z,i
Y g , Z g residuals on GCP and
im
im
im
projection centers (GPS), s xim , s im
y , s X , sY , s Z corresponding to measurements
2011
26
Block adjustment
2011
accuracy a priori, N t number of points measurements on images (point-image pairs),
N g number of GCP and projection centers.

More details about Sigma parameter see in the chapter Brief residuals report on the
Stereopairs tab.
Control, check allows to include a section with ground control and check points residuals:
-----------------------------------------------------------------------------------Ground control point residuals
N
Xm-Xg
Ym-Yg
Zm-Zg
8801
...
9693
0.157
-0.0726
-0.00749
0.173
-0.0175
0.0538
-0.00214
0.0566
mean absolute:
0.0781
RMS:
0.105
maximum:
0.157
number of points (differences):
4 (
4
0.0347
0.0456
0.0726
0.00338
0.00436
0.00749
Xm-Xg
Ym-Yg
Zm-Zg
mean absolute:
0.00
RMS:
0.00
maximum:
0.00
0 (
0
0.00
0.00
0.00
0.00
0.00
0.00
Exy (m)
0.0920
0.114
0.173
4)
projection centers
N
Exy (m)
...
0.00
0.00
0.00
0)
-----------------------------------------------------------------------------------Check points residuals

N
Xm-Xg
Ym-Yg
Zm-Zg
Exy (m)
*212
...
8822
0.989
1.67
-1.02
1.94
-0.0833
-0.0645
0.445
0.105
mean absolute:
0.394
RMS:
0.515
maximum:
1.44
23 (
22
0.528
0.729
1.69
1.57
1.84
3.61
0.750
0.893
1.94
22
22
22)
projection centers
N
Xm-Xg
Ym-Yg
Zm-Zg
Exy (m)
199
...
264
-1.17
0.449
-1.24
1.26
-0.943
0.515
-2.38
1.07
mean absolute:
1.05
RMS:
1.05
maximum:
1.17
3 (
3
0.516
0.518
0.583
35.3
59.1
102.
1.17
1.17
1.26
3)
------------------------------------------------------------------------------------
The option is available only when the option Residuals is marked. The data is output as a
table, where point name and its values are shown:
X = X m X g , Y = Ym Yg , Z = Z m Z g , E xy =
27
(X )2 + (Y )2 ,
PHOTOMOD 5.21
where X m , Ym , Z m coordinates of adjusted (mean) point position, X g ,
Yg , Z g user
defined coordinates.
On stereopairs allows to include a section with adjustment residuals on points measured
on one stereopair at least. The option is available only when the option Residuals is marked.
Depending on the Tie residuals radio button the report contains the following statistical
parameters.
When the option Tie residualsFrom mean value is selected:
X ij = X ij X m , Yij = Yij Ym , Z ij = Z ij Z m , E xy,ij =
(X ) + (Y )
2
ij
ij
where i point number, j stereopair number, X ij , Yij , Z ij coordinates of i point,

calculated from j stereopair; are shown correspondingly in X-X, Y-Y, Z-Z and Exy
columns:
-----------------------------------------------------------------------------------Tie point residuals (from mean)
N
X-Xm
Y-Ym
Z-Zm
mean absolute:
0.103
RMS:
0.134
maximum:
0.510
53 (
122
0.0759
0.0986
0.414
0.149
0.195
0.849
122
122
Exy (m)
0.139
0.166
0.516
122)
-----------------------------------------------------------------------------------Targeted point residuals (from mean)

N
X-Xm
mean absolute:
0.00
RMS:
0.00
maximum:
0.00
0 (
0
Y-Ym
Z-Zm
Exy (m)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0)
-----------------------------------------------------------------------------------Tie point residuals - projection centers (from mean)

N
X-Xm
mean absolute:
0.00
RMS:
0.00
maximum:
0.00
9 (
18
Y-Ym
Z-Zm
2.53e-14
1.07e-13
4.55e-13
0.00
0.00
0.00
18
18
Exy (m)
2.53e-14
1.07e-13
4.55e-13
18)
------------------------------------------------------------------------------------
When the option Tie residuals Between models is selected:
X ijk = X ij X ik , Yijk = Yij Yik , Z ijk = Z ij Z ik , E xy,ijk =
(X ) + (Y )
2
ijk
ijk
where i point number, j and k numbers of stereopairs where i point is measured;

are shown correspondingly in X1-X2, Y1-Y2, Z1-Z2 and Exy columns:
-----------------------------------------------------------------------------------Tie point residuals (between stereopairs)
2011
28
Block adjustment
2011
N
X1-X2
Y1-Y2
Z1-Z2
mean absolute:
0.179
RMS:
0.223
maximum:
0.614
53 (
91
0.0932
0.144
0.543
0.230
0.287
0.851
91
91
Exy (m)
0.225
0.265
0.646
91)
-----------------------------------------------------------------------------------Targeted point residuals (between stereopairs)

N
X1-X2
Y1-Y2
Z1-Z2
mean absolute:
0.00
RMS:
0.00
maximum:
0.00
0 (
0
0.00
0.00
0.00
0.00
0.00
0.00
Exy (m)
0.00
0.00
0.00
0)
-----------------------------------------------------------------------------------Tie point residuals - projection centers (between stereopairs)

N
X1-X2
Y1-Y2
Z1-Z2
mean absolute:
0.00
RMS:
0.00
maximum:
0.00
9 (
9
5.05e-14
1.52e-13
4.55e-13
0.00
0.00
0.00
Exy (m)
5.05e-14
1.52e-13
4.55e-13
9)
------------------------------------------------------------------------------------
The above parameters describe residuals, which will appear during merging of vector objects
and DEM, created on the adjacent stereopairs.
Tie and pass residuals allows to include a section with tie points residuals calculated
separately between stereopairs and strips. Sections Between models and Between strips
will be added to the Tie residuals section of the report. The option is available only when the
option Residuals--on stereopairs is marked.
-----------------------------------------------------------------------------------Tie point residuals (between stereopairs)
N
X1-X2
Y1-Y2
Z1-Z2
mean absolute:
0.179
RMS:
0.223
maximum:
0.614
53 (
91
0.0932
0.144
0.543
0.230
0.287
0.851
91
91
X1-X2
Y1-Y2
Z1-Z2
mean absolute:
0.175
RMS:
0.204
maximum:
0.391
37 (
43
0.0138
0.0198
0.0603
0.253
0.310
0.712
43
43
X1-X2
Y1-Y2
Z1-Z2
mean absolute:
0.191
RMS:
0.255
maximum:
0.614
26 (
26
0.171
0.207
0.543
0.210
0.285
0.851
26
26
Exy (m)
0.225
0.265
0.646
91)
between stereopairs
N
Exy (m)
0.177
0.205
0.391
43)
between strips
N
29
Exy (m)
0.284
0.329
0.646
26)
PHOTOMOD 5.21
------------------------------------------------------------------------------------
Include singles is used only for residuals calculation method from mean value. At that,
the report will contain points measured only on one stereopair and that are not ground
control. The option does not influence on residuals calculated between stereopairs or strips.
When independent stereopairs or independent strips method is used for adjustment, the
residuals on these points are equal to zero.
During bundle adjustment the residuals on these points differ insignificantly from zero. The
difference is defined by residual vertical parallax of relative images orientation. Usually, they
are less than 0.1 of residual vertical parallax value on the terrain.
N
*136
*137
262-264
X-Xm
-0.000306
Y-Ym
Z-Zm
-0.000168
Exy (m)
2.52e-05
0.000349
262-264
1.41e-05
-2.70e-05
0.000144
3.05e-05
------------------------------------------------------------------------------------
At import of external adjustment results, these residuals could be also not equal to zero, the
residual value depends on the adjustment method used.
At import of external adjustment results, you are recommended to mark this option (along
with selected residuals calculating method from mean value) for intermediate control of
consistency of imported relative orientation parameters and points coordinates. In the final
report this option may be turned off if the residuals on single points are negligible.
On images allows to include a section with residuals on images (in mm or pixels).
x = x pr x meas , y = y pr y meas , E xy =
(x )2 + (y )2 ,
where x pr , y pr coordinates of adjusted (mean) point projection position on image, x meas ,
y meas coordinates of point's measurements on image.

-----------------------------------------------------------------------------------Tie point residuals - projection centers (between stereopairs)
x_pr-x_meas
mean absolute:
0.00288
RMS:
0.00579
maximum:
0.0423
190 (
484
y_pr-y_meas
0.00522
0.00669
0.0241
484
Exy (mm)
0.00690
0.00885
0.0458
484)
-----------------------------------------------------------------------------------Targeted point residuals (on images)

N
x_pr-x_meas
mean absolute:
0.000107
RMS:
0.000109
maximum:
0.000131
2 (
4
y_pr-y_meas
0.00306
0.00306
0.00313
4
Exy (mm)
0.00306
0.00307
0.00314
4)
------------------------------------------------------------------------------------
2011
30
Block adjustment
2011
Exterior orientation angles correction allows to include a section with residuals on

exterior orientation angles, if the angles were used in the adjustment.
----------------------------------------------------------------------------------------------------------------------------------------------------------------------Exterior orientation angles residuals
----------------------------------------------------------------------------------------------------------------------------------------------------------------------Ground control
Image
Alpha
264
Omega
Kappa (rad.)
0.0050651339
0.0002582413
0.0008223636
mean absolute:
0.0050651339
RMS:
0.0050651339
maximum:
0.0050651339
number of images (differences):
1 (
1
0.0002582413
0.0002582413
0.0002582413
0.0008223636
0.0008223636
0.0008223636
1)
-----------------------------------------------------------------------------------Check data
Image
Alpha
mean absolute:
0.0000000000
RMS:
0.0000000000
maximum:
0.0000000000
0 (
0
Omega
Kappa (rad.)
0.0000000000
0.0000000000
0.0000000000
0
0.0000000000
0.0000000000
0.0000000000
0)
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------
Details allows to include a section of detailed information about residuals:

----------------------------------------------------------------------------------------------------------------------------------------------------------------------Detailed adjustment residuals information
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------
When the option On stereopairs is marked the following section is added:

-----------------------------------------------------------------------------------Ground control point residuals
203
8801
...
X-Xg
Y-Yg
Z-Zg
Exy (m)
262-264
0.118
0.460
-0.137
0.475
136-138
199-201
0.0171
0.0528
-0.533
-0.462
0.297
0.588
0.533
0.465
X-Xg
Y-Yg
Z-Zg
projection centers
N
Exy (m)
...
-----------------------------------------------------------------------------------Check point residuals
*212
1225
31
X-Xg
Y-Yg
Z-Zg
Exy (m)
197-199
0.711
0.378
0.343
0.805
197-199
0.0863
-1.30
0.165
1.30
PHOTOMOD 5.21
...
262-264
-0.431
-1.16
-0.321
X-Xg
Y-Yg
Z-Zg
197-199
199-201
-0.330
-0.330
0.00443
0.00443
-0.0179
-0.0179
1.24
projection centers
199
...
Exy (m)
0.330
0.330

N
*142
*155
...
X-Xm
Y-Ym
Z-Zm
Exy (m)
197-199
262-264
0.0113
-0.178
0.0992
0.00337
-0.0578
0.293
0.0998
0.178
262-264
264-266
-0.00854
0.0695
0.0123
0.0140
-0.0284
0.0315
0.0150
0.0709
-----------------------------------------------------------------------------------Targeted point residuals (from mean)

N
X-Xm
Y-Ym
Z-Zm
Exy (m)
...
-----------------------------------------------------------------------------------Tie point residuals - projection centers (from mean)
N
136
...
X-Xm
134-136
136-138
0.00
0.00
Y-Ym
0.00
0.00
Z-Zm
0.00
0.00
Exy (m)
0.00
0.00
------------------------------------------------------------------------------------
When the option On images is marked the following section is added:

-----------------------------------------------------------------------------------Tie point residuals (on images)
*136
*137
...
x_pr-x_meas
y_pr-y_meas
Exy (mm)
262
264
7.80e-05
-0.000109
0.00300
-0.00299
0.00300
0.00300
262
264
2.81e-05
-3.70e-05
0.000962
-0.000961
0.000962
0.000962
-----------------------------------------------------------------------------------Targeted point residuals (on images)
*136
sg2345
...
x_pr-x_meas
y_pr-y_meas
Exy (mm)
262
264
7.80e-05
-0.000109
0.00300
-0.00299
0.00300
0.00300
262
264
-0.000110
0.000131
-0.00313
0.00313
0.00314
0.00313
------------------------------------------------------------------------------------
Mark bad points allows to mark in the report the points with residuals over the acceptable
value by * symbol.
2011
32
Block adjustment
2011
-----------------------------------------------------------------------------------limit:
0.200
0.200
0.200
0.200
203
0.116
0.444*
-0.134
0.459*
8801
0.114
-0.496*
0.430*
0.508*
------------------------------------------------------------------------------------
Include acceptable allows to include a section with all residuals for all points according to
options selected on the Report tab.
-----------------------------------------------------------------------------------limit:
0.400
0.400
0.400
0.400
8801
0.114
-0.496*
0.430*
0.508*
9243
0.205
-0.144
-0.167
0.250
9693
-0.455*
0.181
-0.106
0.490*
------------------------------------------------------------------------------------
If the option is off the report will include residual values only on the points where the limit is
exceeded.
-----------------------------------------------------------------------------------limit:
0.400
0.400
0.400
0.400
8801
0.114
-0.496*
0.430*
0.508*
9693
-0.455*
0.181
-0.106
0.490*
------------------------------------------------------------------------------------
Exterior orientation parameters allows to include a section with exterior orientation

parameters of images calculated during adjustment.
----------------------------------------------------------------------------------------------------------------------------------------------------------------------Exterior orientation parameters
----------------------------------------------------------------------------------------------------------------------------------------------------------------------Image: 134
projection center
5907.812105
1964.086641
rotation matrix
-0.0309144854
-0.9994830686
0.9995190724
-0.0308916973
-0.0024328076
0.0089049782
Alpha, Omega, Kappa (rad)
-0.0088256889
-0.0027068461
1496.543074
0.0088255420
0.0027068428
0.9999573905
1.6016930528
-----------------------------------------------------------------------------------Image: 136
projection center
5897.148453
2882.002910
rotation matrix
-0.0067208647
-0.9996325833
0.9999566038
-0.0065489931
-0.0064513995
0.0263022647
Alpha, Omega, Kappa (rad)
-0.0262624684
-0.0066258516
1495.789992
0.0262588731
0.0066258031
0.9996332179
1.5773455105
...
------------------------------------------------------------------------------------
Catalog allows to include a catalog containing coordinates of all points of the block
calculated during adjustment.
----------------------------------------------------------------------------------------------------------------------------------------------------------------------Point catalog
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------
33
PHOTOMOD 5.21
N
Z (m)
203
8801
2400.304
4928.290
2299.701
3454.432
90.356
112.663
3413.020
3600.092
1619.540
2017.084
99.714
91.248
2445.369
3158.506
97.166
84.686
2251.057
3271.090
97.446
95.484
ground control
...
total
5 points
check
*212
1225
...
total
22 points
targeted
*136
sg2345
...
total
2071.706
3998.009
2 points
tie
*137
*138
...
total
2764.921
2111.772
161 points
------------------------------------------------------------------------------------
Print residuals allows to include in catalog the residuals on GCP and check points.
-----------------------------------------------------------------------------------N
X
Y
Z
Exy
Ez (m)
ground control
203
8801
...
total
2400.304
4928.290
2299.701
3454.432
90.356
112.663
0.459
0.508
-0.134
0.430
1619.540
2017.084
99.714
91.248
1.15
1.37
0.714
-0.428
5 points
check
*212
1225
...
3413.020
3600.092
total 22 points
------------------------------------------------------------------------------------
GPS coordinates correction

If projection centers are used as GCP and systematic error on their coordinates is calculated
during adjustment, the report will contain a section with calculated values of the systematic
error. See also the chapter Systematic error compensation.
----------------------------------------------------------------------------------------------------------------------------------------------------------------------GPS coordinates correction for projection centers
----------------------------------------------------------------------------------------------------------------------------------------------------------------------N
Dx
Dy
Dz
Dxy (m)
11169369
11169374
-0.783
-0.0908
1.80
1.75
0.544
0.608
1.96
1.75
mean absolute:
2.43
RMS:
2.93
maximum:
6.28
1.75
1.99
3.29
0.558
0.635
1.06
3.22
3.54
6.54
ground control
...
2011
34
Block adjustment
2011
36 (
36
36
36
mean absolute:
0.00
RMS:
0.00
maximum:
0.00
0 (
0
0.00
0.00
0.00
0.00
0.00
0.00
36)
check
0.00
0.00
0.00
0)
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------
Exterior orientation angles correction allows to include a section with corrections to

exterior orientation angles, if the angles are used during adjustment and the option of
systematic error compensation was on.
----------------------------------------------------------------------------------------------------------------------------------------------------------------------Exterior orientation angles residuals
----------------------------------------------------------------------------------------------------------------------------------------------------------------------Ground control data
Image
Omega
262
264
0.0087633401
0.0087633401
-0.0038763912
-0.0038763912
0.0112346521
0.0112346521
mean absolute:
0.0087633401
RMS:
0.0087633401
maximum:
0.0087633401
5 (
5
0.0038763912
0.0038763912
0.0038763912
0.0112346521
0.0112346521
0.0112346521
...
Phi
Kappa (deg)
5)
-----------------------------------------------------------------------------------Check data
Image
Omega
mean absolute:
0.0000000000
RMS:
0.0000000000
maximum:
0.0000000000
0 (
0
Phi
Kappa (deg)
0.0000000000
0.0000000000
0.0000000000
0
0.0000000000
0.0000000000
0.0000000000
0)
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------
Tie residuals the option defines one of the following calculation methods of tie residuals
between stereopairs.
Between models allows to use pairwise difference between point position, calculated from
each stereopair, which contains the point measurement.
-----------------------------------------------------------------------------------N
X1-X2
Y1-Y2
Z1-Z2
Exy (m)
*142
197-199 - 262-264
-0.116
0.265
-0.429
0.289
------------------------------------------------------------------------------------
From mean value allows to use differences between point position, calculated from each
stereopair, and mean adjusted position.
-----------------------------------------------------------------------------------N
-Xm
Y-Ym
Z-Zm
Exy (m)
35
PHOTOMOD 5.21
*142
197-199
-0.0581
0.133
-0.215
0.145
262-264
0.0581
-0.133
0.215
0.145
------------------------------------------------------------------------------------
Acceptable residuals allows to setup acceptable residuals for block adjustment. These
limits do not influence on adjustment, but are used in the report and for results visualization
on the block scheme. You can setup the same limits for all point types (the All the same
option), or for different limit for each point type (the By point type option).
In first case, you should fill in XY and Z fields in the All panel. In second case, fill in fields in
panels Ground control, Check, Projection centers, Tie, Targeted, Tie-projection centers
for each point type. The limits are setup in coordinate system units of the project.
Scale the field is used to indicate survey scale used for residuals calculation from image
scale to block scale and vice versa. This value is calculated during adjustment automatically,
but you can edit it manually in necessary.
Mean ground sample distance (GSD) this field shows the mean pixel size on the ground
(ground sample distance), which is calculated based on adjustment results, in the working
units of the project.
Residuals print format allows to select format of output of residuals values in the report.
At that two options are available:
Fixed point is used to round off the residuals up to three decimal places:
-----------------------------------------------------------------------------------203
-0.003
0.089
-0.008
0.089
8801
0.076
-0.165
0.026
0.182
------------------------------------------------------------------------------------
Floating point is used to round off the residuals up to three significant figures:
-----------------------------------------------------------------------------------203
-0.00275
0.0888
-0.00847
0.0888
8801
0.0760
-0.165
0.0259
0.182
------------------------------------------------------------------------------------
Model to setup angular system used for output of exterior orientation parameters. Available
options are: alpha, omega, kappa and omega, phi, kappa. See the description of the
parameters in the chapter Appendix. Exterior orientation angular systems.
Units is used to setup angle units in the report. Available options are Radians and
Degrees.
9. Computations
Once parameters are entered, push the
button to start calculations. Each
calculation session uses currently entered parameters. You can see process progress in the
window.
2011
36
Block adjustment
2011
If it takes a long time to complete processing, push the Stop button to cancel it. In this case
you get the result of current iteration. In case of pushing the Cancel button all the adjustment
results will be deleted.
A bad convergence may be caused by errors of the relative orientation, errors in control
points coordinates or incorrect Block adjustment parameters. Sometimes it is due to
incorrect coordinate system choice (left-right). In this case you should change XY axes
direction using the Coordinate system tab in Parameters window (see the chapters
Coordinate systems and Coordinate system tab).
Note. One iteration could be quite long for a big block processing, but the process can
be interrupted only between iterations
If the results of adjustment are acceptable you can save them (by pushing the Save
adjustment results button, see the chapter Finishing the work in PHOTOMOD Solver), then
close PHOTOMOD Solver and pass to PHOTOMOD Core for further project processing (see
user manual Project processing).
You can use the
Restore state icons restore current job state. If adjustment results are
not satisfying you can cancel them and return the block to its initial state using the button
Reset state.
37
PHOTOMOD 5.21
The results of block adjustment are displayed in two graphic tabs for their analysis (see the
chapter Stereopairs tab and Images tab).
10. Stereopairs tab

The Stereopairs tab is used to show results of block adjustment in graphic form. Here you
can see residuals on ground control points and tie residuals between stereopairs.
Stereopairs tab with adjustment results

See the description of Images tab in the chapter Images tab.
The Stereopairs tab uses the same symbols of points, images, stereopairs and strips as the
Source tab, see the chapter Source tab. Additionally to described you can see here white
points the points that with uncalculated coordinates after adjustment (due to lack of point's
measurements on images, for instance), and also points filtered automatically during
adjustment.
Besides, the adjustment residuals are shown graphically (symbols and colors) on the block
scheme, see the chapter Error vectors displaying.
10.1 Error vectors displaying

Errors (residuals) are displayed in the graphic window as vectors of different directions,
shapes and colors and look like the following:
Residuals on control and check points:
XY vector starting in the point directed to the source point position ending with a
small circle
2011
38
Block adjustment
2011
Z vertical vector ending with a small cross-bar

The vector color is green for acceptable residuals and red for exceeded.
Residuals on tie points:
XY diagonal cross with size proportional to X and Y residual components
Z vertical line symmetrical relatively to the point position

The vector color is deep green for acceptable residuals and deep red for those which
exceed residual threshold.
Error vectors
The error vector size is proportional to the residual value and is set up in Residuals window
(see the chapter Residuals window).
10.2 Residuals window

) is used to set parameters of displaying error
The dialog Residuals (opened by the icon
vectors on the block scheme, appeared on the Stereopairs and the Images tabs after the
adjustment procedure. On the Stereopairs tab this window contains the following options.
39
PHOTOMOD 5.21
Residuals window
Residuals dialog includes the following panels and options:
Show defines which error vectors should be displayed:
XY displays error vectors, corresponded to XY errors.
Z displays error vectors, corresponded to Z errors.
Ground displays error vectors, corresponded to ground (control and check) points
Tie displays error vectors, corresponded to ties errors
Excluded at the excluded points.
Points defines which points are to be used to display error vectors for:
All all points
Filter points, selected by filter (see the chapter Points selection by filter).
View scale allows to manage the size of residual vectors on the block scheme. There
are two options: screen and real. If the screen scale is selected, use the lower field to
input the number of units of project coordinate system (meters, feet) in 100 screen pixels.
Push the arrow buttons
in this field to change the screen scale value, to obtain the
best residuals view on the block scheme on the screen. If the real scale is selected, the
size of residual vector (circle) is corresponding to the block scheme scale on the screen.
Therefore, if the scale value in numeric field is x1 the end of residual vector is directed
to the point with coordinates of ground control or check points. That allows to see the
location of ground control points mistakenly measured. At that the vectors are zooming in
and out while the scheme resizing. See also the chapter Error vectors displaying.
Red arrows on the block scheme show the error vectors for point with exceeded residuals,
green arrows with acceptable ones (see the chapter Stereopairs tab and Images tab).
10.3 Brief residuals report on Stereopairs tab

(Brief residuals report) is used to open the Residuals window. This window
The icon
displays RMS, mean absolute and maximum adjustment error values, that is useful for quick
estimation of the results without viewing of a full report. Here you can also see Sigma_0
value, if it is calculated during bundle adjustment.
Sigma value ( 0 ) shows, how adjustment residuals correspond with limits of points
measurements and input GCP (ground control points and projection centers coordinates)
2011
40
Block adjustment
2011
setup a priori. If limits are setup correctly, 0 value approximately equals 1 (30%). If it is
higher, it means that there are errors in points measurements, in initial GCP, or that the limits
are incorrect. If sigma value is less than 1, it also indicates some errors, most likely limits
mismatch. See also the chapter Report tab.
Residuals window
10.4 Tie residuals window

If the tie point is selected, the values of corresponding errors are shown in the Information
window (see the chapter Information window). You can also open the Tie residuals window
by the
41
button to view the graphic interpretation of the errors on the tie point.
PHOTOMOD 5.21
Tie residuals window

In this window the image scale (in pixels) is always used to display errors.
At the upper part of the window there are XY error vectors directed from the adjusted
(average) tie point position to its positions, calculated from each stereopair. The lower picture
shows Z error vectors.
White labels showing errors are shown after click on the end of error vector (with black
cross). The label contains number (name) of stereopair and difference between average
point location and its location on the current stereopair.
Numeric values of appropriate errors are shown in the Information window, see the chapter
Information window on the Stereopairs tab.
Use the slider at the bottom of window for zooming in / out.
2011
42
Block adjustment
2011
10.5 Information window on the Stereopairs tab

point attributes) is used to view the attributes
Information window (opened by the icon
of currently selected point: its type, coordinates and residuals in it.
It looks the same as the Information window opened on the Source tab, see the chapter
Information window on the Source tab.
There is a text field at the upper window part, which shows the name of selected point, if the
only one point is selected.
If the option Editing panel is on the appropriate panel is opened, where you can change
point type (control or check) and vice versa, and also exclude points from the adjustment.
The point type, which it had during adjustment, is coded by point symbol on the block
scheme.
The points type, which it will have during the next adjustment, is indicated by marked check
box near the appropriate type name in this panel. To change this type push one of the
appropriate buttons:
- control
- check
- excluded
You can edit control point coordinates using the window opened by the button
(Edit coordinates), which is used to open the window
for editing coordinates of triangulation points, described in the chapter Editing points.
43
PHOTOMOD 5.21
Information window shows the data for selected point

Panel By ties contains the button
(Edit tie points),
used for tie points editing in special window. See the chapter Editing points.
You can exclude tie point selected from the further adjustment by the button
. Besides,
you may transfer one or several points from control to check and vice versa using the button
.
The rest part of the Information window used to display selected point attributes. It contains
information if the only one point is selected in graphic window.
Geodetic. Shows adjusted geodetic coordinates (adjusted coordinates), source

(real) coordinates entered in PHOTOMOD AT module (reference coordinates),
difference between calculated and source values (residuals), initial weights of
selected point coordinates (reference weights), and values of GPS compensation for
projection centers if they are calculated (GPS compensation).
Ties. Displays information corresponding with ties in the block adjustment. Number of
models number of stereopairs where the point is measured. Residuals (max)
maximum difference between point coordinates calculated from different models. By
models residuals for every model calculated either as difference from the mean
coordinate values (from mean value) or as difference for each pair of models (mutual
residuals)
Type indicates whether the point was ground control, check or rejected during
adjustment, and whether it used as a tie. It also indicates is the point a projection
center, and shows point code (or image code for projection centers)
Images list of images the point was measured on
Stereopairs list of stereopairs the point was measured on

2011
44
Block adjustment
2011
Strips list of strips the point was measured on (on one image at least).
the following way.
Holding Ctrl key pressed you can click the following lines in the window: Tie
Stereopairsstereopair
name,
Images,
Imagesimage
name,
Stereopair,
10.6 Editing points

If necessary to obtain better adjustment results you can edit (re-measure) some doubtful
points, where error vectors exceed the tolerance as shown on Stereopairs tab. To run such
editing, mark the option Editing panel and use the Information window in two following
ways:
change ground (geodetic) point coordinates in appropriate fields on the panel opened by
the button
(Edit coordinates)
Geodetic coordinates editing
change points location on the images where it exists using the window opened by the
button
45
(Edit tie points)
PHOTOMOD 5.21
Selected point editing

The window allows to change points location and re-measure it on some image manually
or automatically using the correlator. See user manual Aerial triangulation for getting more
information about work in Points measurement window.
If you need to edit several points that are selected in the main graphic window, push the
button
to open the list containing points selected.
Double click the points name in the list to open the window for its editing (see user manual
Aerial triangulation) (or use the button Measure or Enter key). To move and select other
points in the list use PgUp, PgDown or arrow keys. Press Esc to close the list.
2011
46
Block adjustment
2011
Editing points from the list

After editing of the coordinates or points location save these measurements and re-adjust the
block to obtain new results.
If the block scheme contains inter-strip points which are not transferred on some strip
(shown by violet sign
on scheme), they should be measured on neighbor images
within the strip (during triangulation or in Editing window opened by the button
, see user manual Aerial Triangulation) to be used in
adjustment with independent strips or stereopairs method.
11. Images tab

Block adjustment results are shown in graphic form also on the Images tab of PHOTOMOD
Solver. This tab displays the residuals on GCP and tie points in image coordinate system.
Tie point residuals are calculated as follows: its geodetic coordinates are measured by
intersection on all adjusted images where the point is located. Then pixel coordinates of this
point are calculated on each image using known geodetic coordinates of the point and
adjusted model. After that the difference of points pixel coordinates calculated and
measured on the image are shown.
Adjustment results are shown on Images tab likewise Stereopairs tab, see the chapter
Stereopairs tab.
47
PHOTOMOD 5.21
Adjustment results on Images tab
Tie points residuals in this window are shown by targeted circle

. Center of the circle
is in the point with adjusted coordinates. Its diameter fits to the maximal residual on the
image (on all images where the point is measured). Residual value in image scale is
re-calculated to meters on terrain using survey scale denominator specified on the Report
tab in the Parameters window, see the chapter Report tab. Residuals exceeding the
tolerance are shown on the block scheme by red symbols, residuals within tolerance by
green. Tolerance limit is set on the same tab Report.
Ground control points errors are displayed in the same way as on the Stereopairs tab, see
the chapter Error vectors displaying.
11.1 Residuals window

Residuals parameters dialogue is calling by the
2011
residuals button on the Images tab.
48
Block adjustment
2011
Residuals visualization parameters in the Images tab

Here you can include/exclude to/from the report tie points residuals (on images) using the tie
option.
The rest options are the same as on the Images tab, see the chapter Residuals window.
11.2 Residuals on image

Residuals on image, to call the window showing the residuals on
Push the button
selected image (by yellow rectangle) and points in their measured positions.
Residuals on image in the Images tab

Each point is shown by the symbol according to its type (the same as on the Source tab, see
the chapter Points displaying). Vector of tie residual is directed from the point to its projection
49
PHOTOMOD 5.21
on the image of its 3D adjusted average position, scaled in accordance with settings in the
Residuals dialogue (see the chapter Residuals window). This vector is ended with small
diagonal cross. For ground control and check points the residual vector is shown in the same
way directed to the projection of user 3D coordinates on the image. For plane and height
points for coordinates not defined by user, adjusted coordinate values are used for residuals
calculating. Ground control point residual vector has small circle on the end.
Residuals vector color depends on their value whether the residual succeed the threshold
or not. The tolerance (in image coordinate system) is equal to user defined tolerance in plane
(XY), multiplied to the survey scale which is input on the Report tab in the Parameters
window, see the chapter Report tab.
The slider below the image is used for image zoom.
11.3 Brief residuals report on the Images tab

(Brief residuals report) is used to open Residuals report window. This
The icon
window displays RMS, mean absolute value and maximum adjustment error values and is
used for quick estimation of the results without viewing of a full report. The window is similar
to Residuals window opened on the Stereopairs tab, see the chapter Brief residuals report
on the Stereopairs tab.
On the Images tab it has the following format:
Ex, Ey, Ez, Exy
Ground control points
0.410*, 0.330*, 0.342*, 0.526*
Check
projection centers
0.325*, 0.325*, 0.486*, 0.460*
Ex, Ey, Exy
tie (in mm)
0.006, 0.012, 0.014
targeted (in mm)
0.003, 0.003, 0.004
Sigma_0 = 1.23
Note. When adjusting single airborne image project, this window will not contain tie
points residuals
11.4 Information window on the Images tab

point attributes) is used to view the
The Information window (opened by the icon
attributes of currently selected point: its type, coordinates and residuals in it. In this window
you can change point type (control-check) and vice versa, and also exclude the points from
the adjustment (rejected points). Besides, you can edit point coordinates in the windows
opened
using
respectively
the
button
(Edit
coordinates) for ground control points and the button

(Edit tie points) for tie points. See the chapter Editing points.
2011
50
Block adjustment
2011
Information window on Images tab

This window are the same as on the Information window opened on the Stereopairs tab,
see the chapter Information window on the Stereopairs tab. Just Tie section of informational
panel is different.
In Tie section are displayed: maximal residuals on the images for the point selected,
multiplicity (number of images where the point was measured), and images list, where the
point was measured, in descending order of tie points residual. For each of those images tie
residuals are shown (for average adjusted position), points measurements on the image
itself, as well as coordinates of adjusted point projection on the image. In addition, for ground
control and check points initial geodetic coordinates projection on the image.
the following way.
Holding Ctrl key pressed you can click the following lines in the window: Tieimages,
Tieimagesimage
name,
Images,
Imagesimage
name,
Stereopair,
12. Report creation

To display full process statistics push the
button on the main menu bar. You
can also open and view the last report created in the current PHOTOMOD Solver session by
pushing upper menu button
51
(last report).
PHOTOMOD 5.21
Report window
The report contains the residual values calculated on control, check, tie points, and
projection centers for the whole block and for every point, the catalogue of points
coordinates, images exterior orientation parameters and GPS corrections and corrections for
exterior orientation parameters. Report content is setup on the Report tab in the Parameters
window (see the chapter Report tab).
Report refers to different values as follows:
X, Y, Z - coordinate values calculated from a model (model is a strip for the method of
independent stereo pairs, and a stereo pair for the method of independent stereo-pairs).
X1, X2, Y1, Y2, Z1, Z2 - coordinates calculated from different models;
Xm, Ym, Zm - average coordinate value calculated from all models;
Xg, Yg, Zg - ground (real) coordinates, input by the user for ground control and check points
Dx, Dy, Dz - constant error of GPS device in projection centers coordinates.
Upper panel of the Report window contains the following icons:
- opens a window for report font change
- allows to edit report (duplicated by F4 hot key)
- search in report (duplicated by Ctrl-F, Ctrl-G and F3 hot keys)
- save report to the text file
- save report to PHOTOMOD system resource
2011
52
Block adjustment
2011
- print report
- printing settings.
13. Console window

Pushing the icon
(Console) allows displaying PHOTOMOD standard event log.
14. Saving the block scheme

(Save block scheme) of the graphic PHOTOMOD Solver window is used to
The icon
save the block scheme to the vector (wmf) or raster (bmp) format. When saving to wmf
vector format the entire scheme picture is saved. In case of bmp format the scheme
fragment bounded by current view window is saved. You can insert the scheme picture into
any documents, reports and so on.
15. Finishing the work in PHOTOMOD Solver

Once you got good adjustment results save them (using the
button) and exit
button). After this is done, further processing of the project is

PHOTOMOD Solver (the
possible (creating DTM, vectorization, building orthophotos).
16. Block adjustment by external programs

In some cases it is suitable to adjust the block in external software and import the results
then into PHOTOMOD system.
PHOTOMOD Solver allows to export adjustment data into other formats. Input export
parameters to the window opened by the button
adjustment programs.
53
- export source data for external
PHOTOMOD 5.21
Parameters of initial data export to AeroSys format

You can export the data into two types of formats: AeroSys and PAT-B.
When exporting to AeroSys format, all measurements of PHOTOMOD project will be
exported to AeroSys project (*.aer file), refer also to AeroSys web site:
http://aerosys.aerogeomatics.com. During adjustment in AeroSys the preliminary
approximation on the current adjustment in PHOTOMOD Solver is used.
Data export to PAT-B format is performed for each type of measurements separately.
The Set similar file names option is intended for set up the same file names and paths for
all marked formats as first file has. At that, file extensions will be different in accordance to
selected formats.
Besides measurements (the coordinates of points on the images, geodetic coordinates of
GCP's, the elements of exterior orientation parameters), the a priory accuracy of these
measurements is written to the exported files as well.
During export it is necessary to set apriori RMS of measurement of points on images in the
Measurement accuracy (std. dev.) field. For measurements obtained from the quality
scanned images, acquired by precisely calibrated camera, the value of measurement
2011
54
Block adjustment
2011
accuracy is 0.5 pixel (e.g., 7 microns, i.e. 0.007 mm). For less quality source images, the
value will be coarser, so it is necessary to estimate it in advance.
Accuracies of geodetic coordinates of GCP's and elements of exterior orientation (projection
centers and omega, phi, kappa angles) are set in the catalogs of GCP's and elements of
exterior orientation (see user manual Aerial triangulation). During export they can be
adjusted by settings their weights (default value is 1) in the Weights panel. In the output file,
the accuracy will be divided by the given weight.
An indirect indication of the correctness of assignment accuracy is the closeness of
adjustment RMS to given apriori values, but this is an unreliable criterion. Big errors in the
entered values can lead to the inability to adjust block.
Image measurements scaling field allows to multiply all the measurements on the images
by the entered factor which can be useful, for example, if it is necessary to export the
measurements to microns, in this case it is necessary to set factor of 1000.
After settings the export parameters and pushing the OK button, the source data will be
exported to the selected files.
You can import adjustment results produced by external software using the window opened
by the button
- import adjustment results.
Import adjustment parameters from PAT-B format

When importing from AeroSys project it is enough to select AeroSys project name (file with
*.aer extension) and path in appropriate field.
You can import two types of data from PAT-B format: rotation matrix and angles, using
appropriate radio buttons in import window.
When importing from PAT-B you can also use an additional re-calculations of initial data,
assigned in the Angles panel:
Negate focus
Invert matrix
55
PHOTOMOD 5.21
Note. During import of adjustment results obtained in AeroSys or other external

software, select Cartesian coordinate system for PHOTOMOD project, to avoid plane
and height errors on images projection centers
See PHOTOMOD Core User Manual for detailed export and import procedures description
using external formats on Block adjustment stage.
17. Single aerial image adjustment

In the current PHOTOMOD version single aerial image adjustment is not supported.
18. Adjustment of scanner blocks

PHOTOMOD project which contains scanner images block
PHOTOMOD Solver after its interior orientation and georeferencing.
is
adjusted
in
Scanner images block on Initial tab

Prior to start adjustment select adjustment parameters in the window opened using the
button. The appeared dialogue contains three tabs: Points, Images and
Report.
2011
56
Block adjustment
2011
Parameters for scanner block adjustment on Points tab

The Reference system panel on the first tab shows coordinate system selected for the
project and its parameters, and also is intended for its change if needed. Push the button
to select new coordinate system in appeared window. For more information
on selecting coordinate system see user manual Project creation.
The Points coordinates accuracy panel allows to setup the accuracy of pixel and ground
(geodetic) points coordinates involved to block adjustment:
The Image measurements RMS (pixel) allows to set up a priory accuracy for tie and GCP
measurements on images (if tie points are involved to adjustment);
Plane set up a priory accuracy of GCP plain coordinates (X and Y) measurements
Height set up a priory accuracy of GCP heights measurements (Z coordinate),
weight of which is 1
Units' errors of plain coordinates and heights weight should be input in the same units as for
project coordinate system.
The Use ties in block adjustment option allows to include/exclude tie points to/from the
adjustment process.
On Images tab in the Method panel select the adjustment algorithm:
Rigorous uses parameters of sensor, which acquired space images.
57
PHOTOMOD 5.21
Adjustment parameters on Images tab

Stereo processing panel is designed for configuring parameters of block processing in
stereoblock mode. When the Create stereopairs option is checked, stereopais are
created from each two adjacent images in a strip. Use adjustment result to create
epipolar images box (by default) - allows to compute epipolar images based on
adjustment results. When this option is on, correctly oriented models (with correct stereo)
can be obtained on stereopairs with complex geometry.
Image Independent refinement is performed only using GCP measured on it, without
ties with other images of the block. For independent refinement it is possible to make
alternate refinement into linear and angular elements of exterior orientation (roll, yaw,
pitch). Refinements are polynomials, orders of which are set up by user by or the system
automatically.
Independent refinement may be executed for obtaining of initial approximation for block
adjustment, or could be used as a main procedure of image orientation (when given
image is excluded from block adjustment), or it could be turned off for selected image (or
for all images).
Beside GCP, block adjustment uses ties between images. During adjustment
refinement polynomial orders are input to images exterior orientation elements. Order
of polynomials is determined automatically, or by user. The image could be excluded
from block adjustment.
RPC this mode uses RPC coefficients delivered by the data supplier along with images
2011
58
Block adjustment
2011
Adjustment parameters for RPC algorithm

stereoblock mode. When the Create stereopairs option is checked, stereopais are created
from each two adjacent images in a strip. Use adjustment result to create epipolar
images box (by default) - allows to compute epipolar images based on adjustment results.
When this option is on, correctly oriented models (with correct stereo) can be obtained on
stereopairs with complex geometry.
When using RPC algorithm, Options group in Refinement panel allows to select refinement
type for RPC, which are delivered alone with initial images:
- Auto the refinement type will be detected automatically using the number of ground
control and tie points;
- Bias if there is enough number of points the constant refinement will be applied to
RPC;
- Bias and drift if there is enough number of points the affine refinement will be applied
to RPC, otherwise the constant refinement will be applied if possible;
- None no refinement will be applied to the initial RPC. This option could be used for
GCP or project coordinate system control.
Options in Parameters apriory covariances field determine initial RPC accuracy as well as
coordinate system correctness (i.e. in case of approximation of local coordinate system with
unknown parameters using Topocentric horizontal coordinate system):
- Bias (pixels) RMS of absolute terms of input refinement (in pixels) a priory;
59
PHOTOMOD 5.21
Drift (p.p.m.) RMS of linear coefficients of input refinement a priory; it is dimensionless

units and their value is setup in millionth parts (p.p.m. parts per million).
Generic allows to select model, which is used for adjustment if the project does not
include interior orientation parameters of the following type: parallel-perspective or DLT
(Direct Linear Transformation), models description see in Project creation User Manual.
If RPC coefficients are not supplied with the product or the results of adjustment using
the coefficients are poor, you may adjust the projects using Generic algorithm. Select the
appropriate option in this case in Parameters window on Images tab:
Adjustment parameters for generic algorithm

After that in Parameters panel you can choose the model type: Parallel-perspective or DLT
(direct linear transformation).
If the number of GCP is sufficient you may leave the model selected by default in that case
the algorithms work similarly. If GCP amount is minimal, use default algorithm first and in
case of improper results, use another one.
stereoblock mode. When the Create stereopairs option is checked, stereopais are created
from each two adjacent images in a strip. Use adjustment result to create epipolar
images box (by default) - allows to compute epipolar images based on adjustment results.
2011
60
Block adjustment
2011
When this option is on, correctly oriented models (with correct stereo) can be obtained on
stereopairs with complex geometry.
Apply button applies setup parameters to the image selected in the left window.
Apply to all button applies setup parameters to all images shown in the left window.
If you would not like to apply setup parameters push Cancel button to close Parameters
window without preserving changed parameters.
After OK button pushing input parameters will be checked and if some values are incorrect,
the tab containing wrong field becomes active and the input field red colored. In case of
correct data input, Parameters window will be closed and adjustment process starts.
Note. Inexperienced users are not recommended to change adjustment settings on
Images tab
Report tab is used to select adjustment report parameters and set the acceptable residual
values.
Adjustment parameters on Report tab

Acceptable residuals panel contains fields for input errors threshold for ground control,
check and tie points. You can consider plane and height errors of points location both on the
terrain (in meters) and on the images (in pixels).
61
PHOTOMOD 5.21
Errors on tie points included to the report are calculated using two methods shown in Tie
points errors panel by the following options:
From adjusted location tie point residual included to the report is maximal
deviation from average coordinate values calculated for all models (stereopairs)
Between stereopairs tie point residual included to the report is maximal difference
between tie point coordinates calculated on the different models (stereopairs).
The panel Report allows to include/exclude to/from the report the following:
Report on images, Report on stereopairs, Report on block includes to the report tie
points residuals, calculated on images, stereopairs and on the whole block
Mark exceeding of limits points with residuals exceeding the threshold (indicated in
Acceptable residuals panel) will be marked by asterisk symbol (*) in the report
Exclude acceptable allows to exclude points with acceptable residuals (setup in
Acceptable residuals panel) from the report
Residuals in meters in the pushbroom block contains stereopairs (see chapter Information window on
the Source tab) are the residuals of the intersection solution: coordinates X calc ,Y calc ,Z calc are calculated
by the stereopair and then compared to the catalogue coordinates X cat ,Y cat ,Z cat .
Residuals in the image overlaps of pushbroom block without stereopairs project may be computed the
same way, but in this case these residuals are not a reliable estimation of the block orientation
accuracy: as the basis to height ration is arbitrary, even with rather accurate orientation of each
image, the intersection solution in the overlap area may have large errors, which is not otherwise
significant for the projects of this type, as intersections are not solved for obtaining output products for
such projects.
Residuals in pixels are the resection solution: from the known X, Y, Z coordinates on the ground, x calc ,
y calc coordinates on the image are calculated and compared to measured coordinates x meas , y meas . In
case of pushbroom block without stereopairs, these errors for the given GCP or check point are
displayed in the Residuals on images section in the Information window on the Images tab
2011
62
Block adjustment
2011
Residuals on image in meters are computed from residuals in pixels and GSD of the image.
There is an option of specifying threshold residuals both in pixels and in meters. In this case a point is
considered erroneous if it fails at least one threshold condition.
Points catalog allows to include to the report the list of adjusted triangulation points
coordinates (X, Y, Z). And if the option Show errors is on plane (Exy) and height (Ez) errors
on these points will be included to the report either. See also the chapter Report tab.
After pushing the OK button the input parameters are checked. If they are incorrect the tab
with incorrect value opens and the wrong field become red. If all the parameters are
acceptable the window is closed.
When adjustment parameters have been setup, start the adjustment by pushing the button
, see the chapter Computations.
Adjustment results will be displayed in graphic window on Stereopairs and Images tabs
(see the chapter Stereopairs tab and Images tab).
Results of scanner images block adjustment

Push the button
Create adjustment report, to view the results in Report
window, see the chapter Report creation.
Then you should save adjustment results (by pushing the
button), if they are
acceptable and go to the next stage of project processing (see the chapter Finishing the work
in PHOTOMOD Solver).
63
PHOTOMOD 5.21
Appendix A. Exterior orientation angular systems

The angular exterior orientation parameters are calculated and included to the report as
omega-phi-kappa or alpha-omega-kappa system of angles.
The rotation sequence omega-phi-kappa.
a) Rotation about X-axis at a positive value of - angle.
Z Y
Z
Y
X, X
b) Rotation about Y-axis at a positive value of - angle.
Z Z
Y, Y
X
X
c) Rotation about Z-axis at - angle.
Y
Z, Z Y
X
2011
64
Block adjustment
2011
The rotation sequence alpha-omega-kappa.

d) Rotation about Y-axis at a negative value of - angle.
Z
Y, Y
X
e) Rotation about X-axis at a positive value of - angle.
Z
Z
f)
X, X
Rotation about Z-axis at -angle.
Z, Z Y
65
Y
X
PHOTOMOD 5.21
Appendix B. Recommendations on aerial images blocks adjustment

This appendix contains recommendations elaborated during adjustment of aerial images
blocks in Racurs company. Scanner survey data processing is not considered here. Tools
and methods described are not supposed to be applied to all possible cases of processing of
central projection data. However, the information stated here might be useful for
PHOTOMOD system users.
1. Typical errors
If you use projection centers for adjustment, you should not use local coordinate system
(which considers Earth curvature approximately). You should use Cartesian coordinate
system (if survey scale and accuracy requirements allow) or real coordinate system, where
points have been measured (SK-42, WGS-84 etc.).
When adjusting without projection centers in Cartesian or local coordinate systems, you
should not consider obtained exterior orientation parameters as accurate values in geodetic
coordinate system. It is especially related to local coordinate system, which leads to big
discrepancies either in height and in plane coordinates of projection centers, as well as in
angles of images orientation. Points' coordinates on the ground are calculated correctly at
that.
Typical errors at distortion setup: distortion input with wrong sign (positive or negative),
distortion units are mixed up: microns millimeters.
2. Stages of control
2.1
2.1.1
Interior orientation control

Interior orientation control
See user manual Aerial triangulation

When perform images interior orientation the systematic radially directed errors may occur.
The reason of such errors is not clear, during adjustment their effect is the same as a results
of wrong consideration of distortion.
It should be noted that some errors of interior orientation (focus error, distortion error) may be
detected reliably only during block adjustment using ground control points (GCP) coordinates
and projection centers.
2.2
2.2.1
Relative orientation control

Points recognition control

If tie points are measured automatically, the typical errors are occurred on interstrip ties
measured in shadows and on trees. Time between shooting of two adjacent images is
usually small enough and object shadow does not shift on significant distance. When
surveying two different strips the time difference often allows the shadow to move on
distance which is close to required accuracy of adjustment. It causes systematic shift of
strips relatively to each other, which could be detected only during adjustment with GCP. At
that it will result in systematic error on projection centers, and a random error on GCP.
2011
66
Block adjustment
2011
You should pay attention to trees shadows and heads: their position may be changed due to
wind. In this case, the error is random or systematic within small area.
You should keep in mind that automatic points' filtering does not work in such cases. The
reason is that those erroneous points in small areas cause self-consistent systematic error,
which is impossible to detect without adjustment using big amount of GCP. If there is a
majority of such points, all the rest will be filtered automatically including those measured
correctly.
One more possible source of errors is measuring of points on borders of extended objects. In
this case, correlation coefficient may be big, but accuracy of points' measurements is high
only in direction perpendicular to the border. Accuracy along the border is usually several
times worse. Automatic filtering is also often unhelpful, since those points may be associated
with other erroneous points. Besides, if the border is directed along epipolar line in stereopair
or at some angle to it, it is impossible to detect such an error automatically.
2.2.2
Control of uniform distribution of points measurements on stereopairs

For relative orientation you should have not less than 5 points, each 3 of them should not be
on the same line. It is recommended to measure more points (not less than 6) to be able to
control errors. You should keep in mind that for algorithm of relative orientation measurement
a group of points measured along straight line is equivalent to two points on this line. It is
necessary to consider when measuring points along man-made extended objects: roads,
power lines, etc.
You should avoid situations when there are many points in some small part relatively to the
rest area of stereopair. You should either thin out these points or add more points on the rest
of stereopair.
2.2.3
Control of vertical parallaxes on stereopairs

In some cases, errors of points' recognition may not result in big vertical parallaxes, but in
predetermined wrong angles of relative orientation, for example, when kappa angle is more
than 45 degrees.
2.3
2.3.1
Control of triplets
Control of uniform distribution of points measurements in triplets

For correct triplets creation it is necessary to measure not less than 2 points with distance
between not less than 1/3 of image width. You are recommended to measure not less than 1
point in each of 3 zones: central and 2 outermost.
If it is impossible to meet this condition (for example, if water body covers more than a half of
strip width) use bundle adjustment for this part of block. At that, you should match it to
adjacent parts of the block in the best way and measure more than usual number of strip
ties. You can also measure "not-transferred" points (that are not measured in images overlap
area), in areas where there is no images overlap but it is necessary to provide high accuracy.
You should keep in mind that for images of this area the exterior orientation parameters are
67
PHOTOMOD 5.21
calculated with less accuracy, thus stereo vectorization accuracy is guaranteed only in the
area that contains tie points.
2.3.2
Control in triplets

During errors control you should consider that height errors in triplets are bigger on average
than plane errors. A coefficient is equal to height to basis ratio of survey.
Too big height errors are usually caused by wrong points measurement or by interior
orientation errors, blunders in focus length or distortion input. If you failed to reduce errors,
check relative orientation.
2.3.3
One strip adjustment in free model using independent models method
You can adjust each strip separately (by selecting of sub-block) in free model. Such kind of
adjustment is the same as errors control in triplets. If the strip is drifted too much in height,
you are recommended to check camera parameters (distortion).
It is necessary to control errors both on the Stereopairs and on the Images tabs.
On this stage you can also detect blunders in focus length input, see the chapter Detection of
blunders in camera focus input. PHOTOMOD Solver.
2.4
2.4.1
Control of free model adjustment

Using of splitting on sub-blocks
Most operations described in this section can be carried out not for the full block, but for
individual strips or groups of strips selected as a subblock. Subblock selection can be useful
for two main reasons. The first one: if the block contains many blunders, it is easy to detect
them by splitting on sub-blocks in such a way that each sub-block would contain not more
than one blunder. The second reason: time for sub-block adjustment is less than time for
adjustment of the whole block that is important for big blocks.
2.4.2
Detection of blunders in camera focus input
Adjustment in free model allows to detect blunders in camera focus input. For this purpose
you should input approximate survey basis. If the basis is correct, Scale field on Report tab
in Parameters window will show correct survey scale value after adjustment. If the difference
between calculated height values of triangulation points and projection centers differs
significantly from real one at that, than more likely focal length was input incorrectly.
Note 1. In free model calculated heights of projection centers are close to 0, and of
triangulation points negative and in absolute value are close to survey height.
Note 2. As survey basis is input approximately, you should compare measured and
real height of survey considering accuracy of this approximation. That is why it is
impossible to find out small errors in focus input on stage of adjustment in free model,
for this the adjustment using GCP is required.
2.4.3
2011
Detection of blunders in strips tying
68
Block adjustment
2011
In order to detect blunders in interstrip tie measurements you are recommended to adjust
sub-blocks (groups of strips) using independent strips method in free model.
You should keep in mind that in case of long strips (20 bases) the errors might exceed
calculated ones in 1020 times, that is caused by adjustment method nature. It is
recommended to split on parts strips with length more than 1020 bases when selecting subblock.
To obtain correct error values in meters you should input survey basis correctly. Usually the
accuracy about 3050% is enough. After adjustment you can check if the basis was input
correctly, using Scale field on Report tab in Parameters window, where calculated survey
scale will be shown.
For blocks with strips of moderate length (up to 15 images), you can adjust the whole block
using independent strips method in free model.
2.4.4
Intermediate control of tie measurements blunders
After detection of errors in independent strips method, it is recommended to identify tie

measurements blunders using adjustment by independent stereopairs method prior to bundle
adjustment. On this stage, the errors that are not exceeding 24 thresholds are acceptable.
2.4.5
Blunders detection on interstrip not-transferred points
Before bundle adjustment, it is advisable to check blunders on interstrip points, that are nottransferred in images overlap on one of the strips. For that, you should adjust block by
independent stereopairs method in free model and compare errors values shown on the
Stereopairs and Images tabs.
If errors circles size on the Images tab in some point exceeds errors vectors size on the
Stereopairs tab in several times, you should check if this point is not-transferred in one of
strips, i.e. it is measured in this strip on the only image. Usually the point happens to be
measured incorrectly on this image.
If the measurements are correct, then more likely in triplets corresponding to this point ties
are located just in small overlap area, and you should add some more ties. At that, as a rule
on the Images tab big circles of errors are existing not on separate points but on all points of
some area.
If the discrepancy occurs, but it is not big (1.52 times), the reason of the error may be big
residual parallax on the point on one or several stereopairs.
2.4.6
How to detect errors of tie points measurements
After the whole block checking for blunders, you should adjust the block in free model using
independent stereopairs method and bundle adjustment to find out moderate errors in
measurements of tie points between strips and images.
You are recommended to reach errors less than 75% of tolerance for ties errors and then
perform final adjustment using GCP, because adjusting with GCP leads to errors increasing.
The first adjustment should be performed by independent stereopairs method, as it is more
stable. Then you may alternate independent stereopairs with bundle adjustment. You should
control errors using both the Stereopairs and the Images tabs independently from
adjustment method.
69
PHOTOMOD 5.21
To obtain correct error values in meters you should input survey basis correctly. Usually the
accuracy about 1020% for the basis value is enough.
2.4.7
How to detect errors of automatic tie points measurements
At automatic tie points measurement you may get a situation when obviously erroneous
points are corrected, and maximal (out of tolerance) adjustment error exists on points
measured correctly. One of the reasons may be incorrect work of automatic tie points
measurement algorithm (due to complex terrain, wrong algorithm settings).
In this case, it is recommended to check regularity of points' distribution and errors absence
on closest stereopairs and triplets, and to perform the following validation procedure for each
point starting from points with maximal error.
1. Check point location (measurement) in stereomode thoroughly.
2. Measure one more point near point under test (on a distance not exceeding of about 1/10
of stereopair overlap area). If after adjustment the errors on both points will be of the same
value (direction and size), then more likely the point is measured correctly.
If the situation is caused by irregular points distribution, the error may be within tolerance
after adjustment, and maximal error may move to another point. If the errors have a value
close to accuracy limit for particular input data, you may stop search for error at that.
3. Check a few closest tie interstrip points (all of them or minimum 3 ones regularly
distributed in each stereopair overlap where the point exists, and some more points in a
distance of 13 stereopairs).
4. Check the same way all stereopairs where the point exists, in every of the six standard
zones falling in the triplet overlap or add 1 tie point in each triplet. If the block has changed,
you must re-adjust it, after that the error may appear on really erroneous point.
5. If maximal error value is on already checked points, you should reject from adjustment the
rest (un-checked) points of stereopairs, where the point exists (you may also exclude part of
points on neighbour stereopairs, keeping in mind regular points' distribution). After
adjustment turn on the option allowing to show errors on points rejected from adjustment.
Check points with maximal errors.
If it is impossible to identify erroneous points after all that procedures, you should check
other possible sources of errors (wrong interior orientation, deformations of some particular
image of the block, adverse choice of objects for points' measurements). In some cases a
good way to reduce an error is to remove all points on several adjacent stereopairs and to
re-measure points in this area manually.
2.5
2.5.1
Control of adjustment of georeferenced blocks

How to detect a reference to wrong coordinate system
If the adjustment of georeferenced block fails, you should check coordinate system selection.
1. Check if it is correct selection of left-handed or right-handed coordinate system. You can
perform it experimentally, by adjusting the block in Cartesian right-handed and Cartesian lefthanded systems. You can also identify a type of used coordinate system by comparing
2011
70
Block adjustment
2011
coordinates of 3 GCP, that are clearly not lie along a straight line, with their location on block
scheme (before adjustment or in free model).
2. Check if the format of points coordinates matches coordinate system selected (is it with
zone number or not). Make sure that there are not input mistakes in coordinates (odd
numbers, skipped numbers, wrong zone number, etc.). Check if the block is adjusted in
Cartesian coordinate system.
You can use independent strips method to speed up verifying.
2.5.2
Detection of blunders in GCP measurements
To detect blunders in GCP measurements you should select a few more reliable GCP and
adjust a block using the rest of points as check. If geometry of block allows you can leave 3
points located as far as possible from each other and that are clearly not lie along a straight
line (desired shape close to equilateral triangle).
On stage of blunders detection you can use Cartesian coordinate system.
If the error value is big enough it is helpful to show errors in real scale with zoom 1, in this
case the end of vector of plane error shows on block scheme real point location with preset
geodetic coordinates. In practice in some cases it turned out that in the list of catalogue
numbers of points group or one of point's coordinates are mixed up.
It would be also helpful to look at suggested location on image of erroneous point. For that
you should mark a point as a check, adjust a block and visualize vector of GCP error on the
point in the Edit point window using the button
Show GCP error. In practice errors
occur due to wrong objects recognition (adjacent power lines pillars, another corner of fence,
etc.).
2.5.3
Selection from several adjacent GCP, if their error' vectors have opposite
direction
If there are two or more points located close to each other, where error vectors are
oppositely directed, you can transfer the points to check and adjust the block. Most likely that
the erroneous point is a point with maximal error, however in some cases it might not be like
that (if for instance, the points are on the edge of block, or the error is caused by other GCP
or tie points).
You can also transfer these points to check one after another, and assign them small values
of weights (0.10.01).
2.5.4
Control of GCP located on one stereopair
If four or more GCP with blunders are located on one stereopair (could be recognized there),
it is helpful to select sub-block containing just this stereopair and analyze it. In this case, the
least number of factors influences on adjustment errors.
2.5.5
Control of adjustment with projection centers
For evaluation of adjustment by projection centers it is useful to perform the following.

Transfer GCP to check points on the area covered by images with known projection centers
coordinates. Make sure that there is no big systematic error.
71
PHOTOMOD 5.21
Big systematic error in height is usually caused by incorrect input of camera focal length or
by selection of wrong coordinate system (wrong value of local plane scale of cartographic
projection).
Big systematic error in height which depends from a distance between the point and block
center, as well as big plane errors directed radially from block center, are usually the result of
improper choice of coordinate system (Cartesian geodetic), errors of interior orientation
(distortion).
2.5.6
Detection of systematic errors of interior orientation
If case of areal survey with big enough amount of GCP (approximately 10 and more) you can
assign small weight of equation on GCP (0.010.1) and adjust then without projection
centers. Systematic error on GCP causes the height errors vectors in block center are
directed oppositely to the vectors on block edges. That may indicate incorrect selection of
coordinate system (inaccurate consideration of the Earth curvature) or errors of interior
orientation (unaccounted distortion).
262-270
197-205
134-142
Vectors of height errors when the Earth curvature is considered improperly or if the distortion
is assigned incorrectly
If systematic error looks in such a way that vectors of plane errors show block squeezing in
one direction and block straining in perpendicular direction, this may be a result of
unaccounted stress/strain of images in one direction.
262-270
197-205
134-142
Errors of stress/strain
2011
72
Block adjustment
2011
You should keep in mind that if you use method of independent strips and independent
stereopairs when GCP weight is considerably reduced (approximately up to 0.01 and less)
then your block might be squeezed (error vectors in plane are directed outward from block
center). It is caused by mathematic properties of these adjustment methods. Bundle
adjustment has no such a property.
2.5.7
Bundle adjustment and adjustment import: errors on points measured on one

stereopair
When using bundle adjustment or importing adjustment results from external software you
should check adjustment errors with marked options Tie residuals from mean value and
Include in the report Residuals on stereopairs include singles, see the chapter
Report tab.
When using bundle adjustment for a survey performed by long focus camera or with big
image overlap (more than 65%) you should consider the following.
Focus enlarging and image overlap increasing result in reducing of basis to survey height
ratio. With this ratio reduction, and adjustment method set to independent stereopairs, height
errors on GCPs increase; on the contrary, when bundle adjustment is used, these errors
decrease (this is the mathematical feature of the algorithm). One can regard it as if height
component of the error on GCPs were transferred to the height component of the error on tie
in these points in bundle adjustment.
That is why height errors on GCP are mainly appeared not in the report chapter devoted to
GCP errors (or GCP errors vectors on block scheme), but in the chapter devoted to errors on
tie points from mean (or shown by corresponding errors vectors).
It should be noted that if using an independent stereopairs method the tie error on points that
are not GCP and measured on the only stereopair, always equals to zero by the
construction. For a bundle adjustment method this errors may differ from zero on a value
depending from vertical parallax residual and difference of distances to projection centers of
two stereopair images on these points. This error is neglible for standard aerial survey with
no blunders in data.
2.6
2.6.1
Control during points filtering in PHOTOMOD Solver

Irregularity of points' distribution control
After points filtering, make sure that there is no stereopairs and/or triplets with irregular
points' distribution.
3. Bundle adjustment specific features

Adjusted heights of interstrip tie points may differ at adjustment using independent strips
method and bundle adjustment.
After bundle adjustment on Stereopairs tab the height error may be shown on such points,
at that deviation from "mean" adjusted position in height from different stereopairs may have
the same direction, that is impossible for independent stereopairs method. The reason is
parallaxes on interstrip ties on stereopairs between images lying on neighbor strips.
73
PHOTOMOD 5.21
If using independent stereopairs method these errors appear as tie errors in plane, and in
bundle adjustment they may transform into height errors.
4. Interpretation of errors shown in PHOTOMOD Solver on

Stereopairs and Images tabs
Errors shown on the Stereopairs tab describe the accuracy, which could be reached, on
further processing during stereo vectorization (including automatic operations performed by
correlator: TIN, DEM building). Errors in triplets and interstrip errors are shown as a residual
during connecting of vector objects created on adjacent stereo pairs.
When using recommended scheme of points measuring in triplets (on triplets' edges and
centers), the residuals of vectors connecting between adjacent stereo pairs are practically
equal to tie errors on triplet points. Errors of interstrip connection are also approximately
equal to errors on interstrip tie points, if they are located on distance of 0.52 bases and are
not lying on the same line as much as interstrip overlap allows.
You should keep in mind that in areas without tie points, discrepancy between vectors may
be greater (due to the degradation of actual triangulation accuracy in the area) by the amount
determined by the points geometry.
Errors shown on the Images tab describe accuracy of orthophoto creation using either
precise external DEM (DEM accuracy is higher than block adjustment accuracy), or DEM
created from block adjustment results (in the first case the average accuracy is expected 1.5
times higher).
Plane and height error values shown on the Stereopairs tab or in the report influence on
orthophoto in so far as they lead to errors on images, although they also influence on
accuracy of created relief model (DEM). At that the value of height error of tie point has less
influence on orthophoto accuracy than error in plane. It should be multiplied by coefficient
that is equal to ratio of distance between point position on image used for orthophoto and
principal point of this image to focal length. This coefficient is always less than one.
5. Processing of survey with big images overlap

The main feature of a survey with big images overlap (more than 70%) is that survey basis to
height ratio is small for stereopairs, composed by neighbour images of strip. Therefore the
value of height error on points measured on these stereopairs exceeds plane errors in
several times. However at that 100% of block area is covered by overlaps of three and even
more images of a strip. In stereopairs composed by images taken as next nearest (one per 2,
3, depending on overlap size), the survey basis to height ratio is big enough like during usual
survey with 60% overlap, that allows to obtain small height error.
The best way of processing of such images blocks depends on kind of output production.
If the end production is orthophoto creation using exterior DEM, the best method will be
processing of the whole block. It is recommended to measure every point (tie, ground, check)
on all images of a strip where it exists, and perform bundle adjustment then. In adjustment
report you should consider only errors on images and errors on GCP and check points
(shown on the Images tab). Tie points errors on stereopairs (shown on the Stereopairs tab)
may be big at that, however they are not influence on orthophoto accuracy, so they should
not be considered.
If you need to produce 3D vector objects or DEM, you are recommended to reduce number
of images in such a way that images overlap in resulting block will be close to 60%, and then
2011
74
Block adjustment
2011
process this block in a usual way. Theoretically, such a reducing is possible when overlap is
more than 70%, but in practice, it is available when overlap is not less than 7580%.
You can evaluate approximately a reducing coefficient as an integer part of an expression
(40/(100k))1, where k images overlap in percent. If the result obtained is 0 it means that
you have to take all of images, 1 next nearest, 2 next third, etc.
If the end production includes both orthophoto and vectors/DEMs, you may use the second
method (as it is less labour-intensive). Another way is to create two blocks: whole and
reduced, then adjust the first one as described above, and import adjustment results into the
second block. The first block will be used for orthophoto creation, and the second one for
stereo vectorization. This method is more laborious, but allows to produce more accurate
orthophoto map.
6. Recommendations on long-focus camera survey

If the end production is orthophoto mosaic, you can execute long-focus camera survey using
standard workflows.
For stereo vectorization and DEM creation long-focus camera survey is not recommended
due to big errors in height obtained during its processing.
Some digital cameras acquire rectangle images with considerable difference between width
and height. If you perform a survey using long-focus camera you should turn the camera in
such a way that long side of image will be oriented along the flight with standard image
overlap (60%), that will allow to reduce height errors. In this case survey basis-height ratio
will be maximal, therefore height errors during block adjustment will be minimal. At that
number of strips will increase.
If there is not GCP enough in the project you are recommended to setup your equipment to
acquire strips with overlap more than usual 20% (3040% for instance), to provide more
strict connection of strips (by measuring interstrip ties in zigzag).
7. Tie points measuring in forested areas, in areas with bad stereo

effect. Tie points measuring between strips acquired in different
seasons
Sometimes it is necessary to process images blocks, consisting of strips acquired in different
seasons, or images of terrain, which does not contain enough number of objects suitable for
stereo viewing and therefore not good for tie points measurements (dense forest areas, for
instance). At that you face a problem of tie point measurement, which could be resolved in
similar ways but with some peculiar features.
The only resolution is usually tie points measuring on artificial objects: power lines pillars,
buildings, roads.
Suppose we need to tie strips of images acquired in different seasons, and each of them is
suitable for stereo processing. Then in those areas where there is interstrip stereo, you may
measure interstrip tie point on the object in each strip independently. At that when measuring
points on the ground level you should consider possible height difference of grass covering
object's base.
75
PHOTOMOD 5.21
Tie points measuring on trees or trees' shadows in dens forest areas is not recommended,
excluding those cases when this area could not be excluded from images block and
estimated residual between neighbor strips will exceed the error caused by wrong point
recognition (corresponding points recognition error and unknown shift of the object due to
tree swinging). Adjustment accuracy in this area and neighbor ones, defined by block
configuration, is worsening correspondingly.
8. Tie points measuring on lengthy objects boundaries

Considering current limitations of adjustment procedure in PHOTOMOD system the common
recommendation for points measuring on lengthy objects boundaries is as follows: if the
boundary or its vicinity has no any features along the boundary, you should not place and
measure tie point there. Features examples: boundaries bends, other objects crossing the
border, common irregular background allowing to perform reliable correlation along the
border.
Lengthy objects boundaries are not only boundaries of objects but also roads (including ruts),
arable lands (consisting of parallel furrows) and other similar objects.
2011
76

PHOTOMOD. Block Adjustment PDF

Uploaded by

Copyright:

Available Formats

PHOTOMOD. Block Adjustment PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

PHOTOMOD. Block Adjustment PDF

Uploaded by

Copyright:

Available Formats

Digital Photogrammetric System

PHOTOMOD Solver. Module of the block adjustment ....................................................... 5

Installation and starting ........................................................................................................ 5

Adjustment of projects of different types ........................................................................... 5

Coordinate systems .............................................................................................................. 6

Free model .............................................................................................................................. 7

Source window ...................................................................................................................... 7

Points displaying ...................................................................................................................... 8

View parameters ...................................................................................................................... 9

Points selection by filter ......................................................................................................... 10

Points window ........................................................................................................................ 12

Images window ...................................................................................................................... 13

Stereopairs window ............................................................................................................... 13

Information window on the Source tab .................................................................................. 13

Images, stereopairs and strips displaying ............................................................................. 15

Cursor modes ........................................................................................................................ 15

Graphic windows toolbar ....................................................................................................... 16

Methods of block adjustment ............................................................................................. 17

Parameters window ............................................................................................................. 18

Coordinate system tab ........................................................................................................... 18

Report tab .............................................................................................................................. 25

Stereopairs tab ..................................................................................................................... 38

Error vectors displaying ......................................................................................................... 38

Residuals window .................................................................................................................. 39

Brief residuals report on Stereopairs tab ............................................................................... 40

Tie residuals window ............................................................................................................. 41

Information window on the Stereopairs tab ........................................................................... 43

Editing points ......................................................................................................................... 45

Images tab ............................................................................................................................ 47

Residuals window .................................................................................................................. 48

Residuals on image ............................................................................................................... 49

Brief residuals report on the Images tab................................................................................ 50

Information window on the Images tab .................................................................................. 50

Report creation .................................................................................................................... 51

Console window .................................................................................................................. 53

Saving the block scheme .................................................................................................... 53

Finishing the work in PHOTOMOD Solver ........................................................................ 53

Block adjustment by external programs ........................................................................... 53

Single aerial image adjustment .......................................................................................... 56

Adjustment of scanner blocks ........................................................................................... 56

Appendix A. Exterior orientation angular systems ......................................................................... 64

Typical errors ....................................................................................................................... 66

Stages of control ................................................................................................................. 66

2.4.2 Detection of blunders in camera focus input ......................................................................... 68

Bundle adjustment specific features ................................................................................. 73

Interpretation of errors shown in PHOTOMOD Solver on Stereopairs and Images tabs

Processing of survey with big images overlap ................................................................ 74

Recommendations on long-focus camera survey ........................................................... 75

Tie points measuring on lengthy objects boundaries ..................................................... 76

1. PHOTOMOD Solver. Module of the block adjustment

2. Installation and starting

3. Adjustment of projects of different types

Project type selection

starts the process of computing of exterior orientation parameters and

Initial block state on the Source tab

XYZ points double triangle

rejected from adjustment white color

projection centers small black square

measured on at least one stereopair small diagonal cross