Research on Human Abnormal Behavior Detection Based on Optical Flow Energy
Luo Chao-y · Computer & Network · 2014
This paper studies and compares three commonly-used kinds of abnormal behavior detection algorithms, analyzes their research status, introduces the research progress of optical flow energy used in abnormal behavior detection technology, proposes a human abnormal behavior detection algorithm based on regional optical flow energy. The Lucas-Kanade algorithm is used to calculate the regional optical flow characteristics and realize the severe abnormal behavior detection in the video such as fighting. The simulation test shows that the algorithm has such features as higher recognition rate and lower false detecting rate and presents certain research significance.