Video-based bbnormal behaviour detection in smartsurveillance systems
Othman Omran Khalifa, Hazwani Abdul Khodir, Noreha Abdul Malik · The International Islamic University Malaysia Repository (The International Islamic University Malaysia) · 2020
Due to increasing demand for Security, the instant detection of abnormal behaior in video surveillance systems becomes a critical issue in a smart surveillance system. The currently applied semiautomatic systems mainly depend on human intervention to detect the abnormal activities and suspicious human behaviours from video context. Due to these limitations, it has become an urgent need for intelligence systems to avoid the very slow response and reduce the human observer and interventions . In this paper, a method that can trace abnormalities of human behaviour from video is presented. Techniques related to bounding box measurements and descriptions for behaviour representation were used. Moreover, the performance evaluation of the proposed method is presented.