Performance Analysis of Branching Particle Filter for Moving Object Tracking and Anomaly Detection
In this paper, we demonstrate video feature representation which isimportant aspect for video anomaly detection (VAD).Research is going on to finding to perform VAD streams accurately with an acceptable false alarm rate. But due to change in nature and humanity and high space-time complexity, it is very difficult with huge video data. Abnormal event detection is practical applications of video surveillance. With problems for VAD systems on practical part: limited labeled data, ambiguous definition of “abnormal” and expensive feature engineering steps. Here, we introduce a unified detection framework to deal with these challenges using different filtering method.Here, we compare performance of Kalman filter, particle filter and branching filter. The experimental result shows that branching filter having more accuracy as compare to other filters.