Scribd Logo
Upload

Welcome to Scribd!

  • 4 views

    MINI

    Uploaded by

    Shubham Pawar

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd

    MINI

    Uploaded by

    Shubham Pawar
    4 views26 pages

    Original Title

    MINI PPT

    Copyright

    © © All Rights Reserved

    Available Formats

    PPTX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    4 views26 pages

    MINI

    Uploaded by

    Shubham Pawar

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    of 26

    Internal Guide : Prof.

    Satish Kuchiwale
    OUR TEAM MEMBERS

    RAJAN CHAUHAN 05
    LOKESH KAPSE 19
    SHUBHAM PAWAR 34
    TABLE OF CONTENT
    INTRODUCTION
    REVIEW OF LITERATURE
    INTRODUCTION
    ➔ The monitoring of behaviour, activities, or other changing information, usually of people
    or places for the purpose of influencing, managing, directing, or protecting them is
    termed as surveillance.
    ➔ The surveillance methods can include observation from a distance by means of
    electronic equipment such as closed-circuit television (CCTV) cameras, or interception
    of electronically transmitted information such as Internet traffic or phone calls, and it can
    include simple, relatively low-technology methods such as human intelligence agents
    and postal interception.
    ➔ Many organizations and people are deploying video surveillance systems at their
    locations with Closed Circuit TV (CCTV) cameras for better security. The captured video
    data is useful to prevent the threats before the crime actually happens. These videos
    also become good forensic evidence to identify criminals after the occurrence of crime.
    PROBLEM STATEMENT

    ⮚ Traditionally, the video feed from CCTV cameras is monitored by human operators. These
    operators monitor multiple screens at a time searching for anomalous activities.
    ⮚ This is an expensive and inefficient way of monitoring. The process is expensive because the
    operators are on a payroll of the organization and inefficient because humans are prone to
    errors. A human operator cannot efficiently monitor multiple screens simultaneously.
    ⮚ Also, the concentration of an operator will reduce drastically as time passes. One of the
    methods to cope with this problem is to use automated video surveillance systems (video
    analytics) instead of human operators. Such a system can monitor multiple screens
    simultaneously without the disadvantage of dropping concentration.
    OBJECTIVE

    Following are the Main Objectives of Our Project

    ➔ To detect Criminal Activities.


    ➔ To Monitor and manage Suspicious Activities.
    ➔ To Avoid Criminal & Suspicious Activities from Taking Place as much as Possible.
    ➔ To enhance the safety and security in residential and Commercial areas.
    ➔ To Improve the Emergency Responses.
    ➔ To Provide High Security.
    SCOPE

    ➢ In the Public Places.


    ➢ In the School and Colleges for Student Safety.
    ➢ In the Residential Areas.
    ➢ In the Offices and Corporate Sectors.
    REVIEW OF LITERATURE

    ● Automated Real-Time Detection of Potentially Suspicious Behavior in Public Transport


    Areas.
    Elhamod and Levine proposed a complete semantics-based solution to the behavior
    detection problem that addresses the whole process from pixel to behavior level.
    Furthermore, the processing is achieved in real time. In this method it is assumed that the
    foreground blobs are extracted in each frame using a conventional background subtraction
    method. These blobs represent the silhouettes of animate (e.g., people) and inanimate (e.g.,
    luggage) objects in the scene, which are the semantic entities associated with the events
    described. However, in practice, it is noted that a single blob will often represent multiple
    objects occluding or standing next to each other. After all blobs have been extracted,
    inferences are made to segment, track, and classify the objects that they represent. Finally,
    the anomalous events must be labeled. Thus a 3-D object level information is obtained by
    detecting and tracking people and objects using blob-matching techniques. Color
    histograms are used for blob matching. Kalman filter is used for tracking purposes. Then,
    the activities are classified into various predefined suspicious activities using a semantics
    based approach. The semantics-based approach replaces the need for training with a more
    straightforward process based on human reasoning and logic.
    REVIEW OF LITERATURE
    ● Fast and Robust Algorithm of Tracking Multiple Moving Objects for Intelligent Video
    Surveillance Systems.
    Kim deals with detecting and tracking multiple moving objects through a single
    camera. The proposed method uses red-green-blue (RGB) color background modeling to
    extract moving regions. Blob labeling is used to group moving objects. This method is
    suitable for the real-time surveillance system because of the fast computation and is
    robust against the environmental influences. To detect the moving objects, RGB BM
    with a new sensitivity parameter was employed to extract moving regions, morphology
    schemes to eliminate noises, and blob-labeling to group the moving objects. To track the
    groups of the moving objects, a tracking algorithm was proposed consisting of the
    prediction of the position of each group, the recognition of the same group, and the
    identification of newly appearing and disappearing groups.
    HARDWARE & SOFTWARE REQUIREMENTS

    HARDWARE SOFTWARE
    ● Windows 7 ● Visual studio Code

    ● Webcam ● Pycharm / Python IDLE

    ● RAM 2GB and above Libraries used

    ● Intel core i3 processor ● Keras module

    ● Tensorflow

    ● OpenCV
    DATA FLOW DIAGRAM

    LEVEL 0
    DATA FLOW DIAGRAM

    LEVEL 1
    AGILE MODEL

    Agile process model" refers to a software


    development approach based on iterative
    development. Agile methods break tasks into
    smaller iterations, or parts do not directly
    involve long term planning. The project scope
    and requirements are laid down at the beginning
    of the development process. Plans regarding the
    number of iterations, the duration and the scope
    of each iteration are clearly defined in advance.
    DATA ANALYSIS

    Data processing means carrying out


    operations on data,especially by a
    computer, to retrieve, transform, or
    Classify information. The Processing
    stage is where raw data is transformed
    into meaningful information.
    SYSTEM DESIGN ARCHITECTURE
    METHODOLOGY

    This is our methodology we are going to follow the steps to perform the video
    surveillance and suspicious movement detection.

    1. The configure of the camera settings are such as resolution, frame rate, and field of view, to
    capture the required footage.
    2. The cameras will then record video footage continuously or at specific intervals, depending on the
    settings. This footage can be stored locally or on a remote server.
    3. To detect suspicious movements in real-time, the video footage is analyzed using computer vision
    algorithms, which can recognize specific patterns, such as the movement of people, objects, and
    vehicles.
    4. When suspicious movement is detected, an alert system can be activated to notify the security
    team, who can then respond accordingly.
    5. After an event, the video footage can be reviewed and analyzed to identify the cause of the
    suspicious movement, as well as to improve future surveillance methods.
    CONCLUSION
    The work proposed a suspicious activity detection from the surveillance video using convolutional
    neural network for feature extraction and a discriminative deep belief network for action classification.
    Compared with the previous works, the proposed approach achieves better classification by deep-
    learning-based model. After humans are detected using a background subtraction method, seven
    frames are selected in which the area of the bounding boxes calculated for the humans detected are
    larger among all.Then this output is fed to a DDBN which is used to classify the recognized actions into
    normal and suspicious actions by training the system using semi supervised learning method. The
    deep learning model ensures more accuracy and lesser false positives.
    FUTURE SCOPE

    It have tremendous scope for research and development. Intelligent techniques, esp.
    Neural and SVM are emerging as indispensable techniques for behaviour analysis.
    Semantic analysis for Activity Detection is also taking hold due to its relative simplicity.
    To make the system fully automatic and robust and to overcome the gaps and
    limitations like information loss, occlusion, noise, non-rigid motions etc., multi-view
    tracking and can be implemented using multiple cameras. Multi view tracking has the
    advantage over single view tracking because of wide range and multiple view
    captures,with different viewing angles for the objects to be tracked. Scope of research
    lies in improving accuracy of such systems along with cost optimization, and making
    them more reachable and applicable for society.
    THANK U !
    THANKS
    Do you have any questions?

    [email protected]
    +91 620 421 838
    yourcompany.com

    Please keep this slide for attribution

    CREDITS: This presentation template was created by Slidesgo,


    including icons by Flaticon, and infographics & images by Freepik

    You might also like