Scribd Logo
Upload

Welcome to Scribd!

  • 15 views

    Naveen Kumar HSDS18D005

    Uploaded by

    kavipriya
    The document provides a progress report from a doctoral student for their 4th committee meeting. It includes details of publications, conferences and seminars attended. It also provides an abstract, introduction and proposed methodology for a project on detecting abnormal human behavior in crowded environments using video surveillance. The methodology involves background subtraction, object detection, speed measurement, analytics and future frame prediction. Block diagrams and algorithms are presented to explain the approach.

    Copyright:

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

    Naveen Kumar HSDS18D005

    Uploaded by

    kavipriya
    15 views18 pages
    The document provides a progress report from a doctoral student for their 4th committee meeting. It includes details of publications, conferences and seminars attended. It also provides an abstract, introduction and proposed methodology for a project on detecting abnormal human behavior in crowded environments using video surveillance. The methodology involves background subtraction, object detection, speed measurement, analytics and future frame prediction. Block diagrams and algorithms are presented to explain the approach.

    Copyright

    © © All Rights Reserved

    Available Formats

    PPTX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document provides a progress report from a doctoral student for their 4th committee meeting. It includes details of publications, conferences and seminars attended. It also provides an abstract, introduction and proposed methodology for a project on detecting abnormal human behavior in crowded environments using video surveillance. The methodology involves background subtraction, object detection, speed measurement, analytics and future frame prediction. Block diagrams and algorithms are presented to explain the approach.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    15 views18 pages

    Naveen Kumar HSDS18D005

    Uploaded by

    kavipriya
    The document provides a progress report from a doctoral student for their 4th committee meeting. It includes details of publications, conferences and seminars attended. It also provides an abstract, introduction and proposed methodology for a project on detecting abnormal human behavior in crowded environments using video surveillance. The methodology involves background subtraction, object detection, speed measurement, analytics and future frame prediction. Block diagrams and algorithms are presented to explain the approach.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    You are on page 1of 18

    4 TH DOCTORAL COMMITTEE

    MEETING
    By
    S. Naveen Kumar, HSCS18D005
    Under the Guidance of

    Dr. R. Shobarani, Associate Professor,


    Dr. M.G.R. Educational & Research Institute
    Date: 01/03/2022
    PROGRESS REPORT

    • Published a paper in Parishodh Journal (UGC Care) ISSN NO:2347-6648 titled Edge
    Computing Based Distributed Video Analytics Using Federated Learning Technique.
    • Attended a webinar on “Artificial Intelligence and its Application” organized by Shri
    Krishnaswamy College for Women, Chennai
    • Attended 5 days online STTP on “Recent Trends and Applications of Machine Learning
    and Deep Learning in IT” organized by the Department of Information Technology,
    RAIT, Navi, Mumbai
    • Attended a webinar on Artificial Intelligence and its applications conducted by
    Department of Computer Science, Vels Institute of Science, Technology & Advanced
    Studies
    2
    ABSTRACT

    Smart Video Surveillance and Human Mysterious Behavioral Recognition in


    a Crowded Environment

    • The proposed method uses Low rank sparse to track objects with abnormal
    motion and highlight them. This method consists of two steps: background
    subtraction and application of saliency measurement. Two matrices are formed,
    1. Low Rank Matrix
    2. Structured Sparse Outlier.
    • Motion detection – average speed of moving objects calculation, abnormal
    human action detection and recognition of suspicious activity. Further action
    will be predicted based on the detected suspicious activity. 3
    INTRODUCTION

    • Human behavior analysis by computer vision techniques has


    been an interesting issue for researchers.
    • Video Surveillance has been used in many applications
    including elderly care and home nursing etc.
    • Smart video surveillance systems are capable of enhancing
    situational awareness across multiple scales of space and
    time.
    4
    INTRODUCTION

    • Now a days crime rate is increasing on crowded area like railway station,
    bus stand, malls etc.
    • The video surveillance in such places are usually monitored by a set of
    security personals and they cannot monitor more than two videos efficiently.
    • Abnormal event detection is a rapid development task in video analysis,
    which is aimed to distinguish abnormal and normal events in surveillance
    videos.
    • The normal and abnormal events have some similarities, more
    discriminating methods or motion information need to be explored.
    5
    PROBLEM DEFINITION

    • The reason being is the instability incidents that are happening all around
    the world.
    • Therefore, there is a need of a smart surveillance system for intelligent
    monitoring that captures data in real time, transmits, processes and
    understands the information related to those monitored.
    • The video data can be used as a forensic tool for after-crime inspection.
    • Still there is no method that avoids the crime happening by giving alert
    or increasing security especially in malls or other crowded area.
    6
    MOTIVATION

    • Motion and saliency monitoring with video processing can


    provide efficient and accessible means to predict and
    prevent abnormal or criminal activities.
    • Detection of anomalous behavior in surveillance feeds frame
    of unusual activity for threat detection and immediate action
    to prevent any misbehavior.

    7
    PROPOSED SYSTEM

    • In this work we are going to propose abnormal activity detection on video


    surveillance.
    • Robust analysis mechanism to handle background regions and foreground
    analysis.
    • Observed Scene constitute of two terms
    • Low rank matrix (Background)
    • Structural sparse outlier matrix (Foreground)
    • Saliency measurement to dynamically estimate the support of foreground.
    • Saliency gives average motion speed and abnormal speed.
    8
    GENERALIZED EXPERIMENTAL SETUP FOR MOVING
    OBJECT DETECTION

    Start Temporal Differencing


    Tracking • Start video ,|-
    Program • Read first frame and set it -
    as background image and
    rest as current image

    Second segmentation Displacement detection


    First • Estimate center point of each object in a
    • Performs connected component
    Segmentation frame using connected component analysis
    labelling again on pixels that and mark the bouncing box. Repeat the
    • Segments each belongs to the sketch we got in process for all the frames and calculate then
    motion region and prior step. Segments that are change of center point between adjacent
    get rid of any too big will be erased. Those frames. Also detect number of frames in
    regions that are segments are background which object appear. Ratio of displacement
    too small. mislabeled from the true object change and frame count gives the speed of the
    object 9
    METHODOLOGY

    Saliency Prediction
    Video Background and and Video
    Capturing Subtraction Analytics Popup

    Snap or Object Speed


    Frame Detection Measurem
    Separation ent

    10
    ALGORITHM

    • Background subtraction – RPCA (Robust Principle Component Analysis) for


    frame subtraction
    • IMTSL (Incremental Multi-Feature Tensor Subspace Learning) for foreground
    detection.
    • Analytics – DLBT (Deep Learning Based Tracking algorithm)
    • Prediction – Future Frame Prediction

    11
    BLOCK DIAGRAM

    Dynamic Dynamic
    Input Video Frame Separation extraction of back extraction of Fore
    ground image ground image

    Object Thresholding and


    Action detection displacement morphological
    estimation operation

    12
    ADVANTAGES

    • Accuracy of estimation of moving objects is high.


    • Proposed methodology increases accuracy of average and
    high speed motion in surveillance video.

    13
    MODEL OUTPUT

    14
    MODULE FOR SPEED ESTIMATION

    15
    APPLICATIONS

    • Video surveillance
    • Malls
    • Hospitals
    • Park

    16
    REFERENCES

    • S. Mhatre, S. Varma, R. Nikhare, Visual Surveillance Using Absolute Difference Motion Detection,
    International Conference on Technologies for Sustainable Development (ICTSD), pp. 1-5, 2019.
    • Hajer Rahali, Zied Hajaiej, Noureddine Ellouze, Robust Features for Speech Recognition using Temporal
    Filtering Technique in the Presence of Impulsive Noise, International Journal of Image, Graphics and Signal
    Processing (IJIGSP), Vol.6, No.11, pp. 17-24, October 2018.
    • J. Huang, S. Hsu1and, C. Huang, Human Upper Body Posture Recognition and Upper Limbs Motion
    Parameters Estimation, IEEE Signal and Information Processing Association Annual Summit and
    Conference, pp. 1-9, 2015.
    • Dongping. Zhang, Jiao.Xu, Yafei.Lu, Huailiang. Peng, Dynamic Model Behavior Analysis of Small Groups,
    IEEE Conference Based on Article Video Wireless Communications & Signal Processing (WCSP), pp.1 – 6,
    2016.
    • S. Rautaray, A. Agrawal, Real Time Multiple Hand Gesture Recognition System for Human Computer
    Interaction, International Journal of Intelligent Systems and Applications(IJISA), Vol. 4, No. 5, pp. 56-64,
    17
    May 2017.
    Thank You

    18

    You might also like

    pFad - Phonifier reborn

    Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

    Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


    Alternative Proxies:

    Alternative Proxy

    pFad Proxy

    pFad v3 Proxy

    pFad v4 Proxy