UTAR Institutional Repository

Vandalism video analysis employing computer vision technique

Muk, Britney Yuen Kuan (2022) Vandalism video analysis employing computer vision technique. Final Year Project, UTAR.

Download (4Mb) | Preview


    In this advance era of technology, the computer vision technique is involved regularly in surveillance system compare to last decade. This project is carried in the field of image processing to solve and improve the problem of vandalism activity occurred in the town that lead to massive destruction and repair costs, also to protect public and private properties by preventing vandalism. This project is a development of an intelligent surveillance system that can detect vandalism events and the proposed novel method is implemented with the technique of YOLO detection, suspicious characteristic detection, and background changes detection. In the proposed system, the method monitors the changes inside the captured scene. When there is a human enter the scene and there are significant changes, indicating damage, a vandalism event is declared and warning alert will be triggered to scare the vandals off. On the other hand, the characteristics and behaviour of the suspicious vandal will also be monitored. Warning is flagged when the probability of vandal behaviour is exceeding a threshold and the early warning will be given out to prevent the vandalism events. The method is tested on the UCF_Crime dataset with around 50 different videos containing vandalism scenes such as spraying paint, breaking windows, defacing public property and etc

    Item Type: Final Year Project / Dissertation / Thesis (Final Year Project)
    Subjects: Q Science > Q Science (General)
    T Technology > T Technology (General)
    Divisions: Faculty of Information and Communication Technology > Bachelor of Computer Science (Honours)
    Depositing User: ML Main Library
    Date Deposited: 13 Oct 2022 15:40
    Last Modified: 13 Oct 2022 15:40
    URI: http://eprints.utar.edu.my/id/eprint/4639

    Actions (login required)

    View Item