Automatic surveillance analyzer using trajectory and body-based modeling

With the continuous improvements in computer-vision techniques, automatic low-cost video surveillance gradually emerges for consumer applications. Successful trajectory estimation and human-body modeling facilitate the semantic analysis of human activities in video sequences. We propose a fast analyzer for surveillance video, which aims at automatic analysis of human behavior and semantic events. Our analyzer employs visual cues to track moving persons and classify human-body postures from a monocular video. It consists of three processing steps: (1) multi-person detection, (2) persons tracking with trajectory estimation, and (3) posture classification. We show attractive experimental results, highlighting the system efficiency and classification capability.