The Study Of Vision-Based Human Behavior Analysis Algorithm

The Study of Vision based Human Action Recognition is one of hot researches in the computer vision field. Its application in intelligent surveillance, human movement analysis, advanced user interface, virtual reality and video coding and transmission areas is very promising. Due to the complexity of human movement itself and the diversity of outside environments, make human action recognition having certain challenging. Therefore, research on vision-based human movement analysis has important theoretical significance and practical value.The paper launches the research around the related problems of vision-based human action recognition, mainly including moving object extraction, moving object tracking, feature. extraction and algorithm of human action recognition four aspects.In the aspect of moving target detection, the Surendra background subtraction and four frames difference algorithm, with using dynamic threshold, are used in the motion objects. Finally, after image processing, moving target can be detected completely.In the part of moving target tracking, on the basis of analyzing of existing tracking method, the blob tracking based on improved particle filter tracking algorithm is proposed with using OpenCV technology. Improved particle filter algorithm means that each particle is optimized by Mean Shift method to improve the tracking speed. Experimental results show that the new algorithm avoids the problem of shade from other objects and the moving objects lose.In terms of feature extraction, the paper extracts seven invariant moments of human motion using improved Hu moment. It gets the seven improved Hu moment by extracting the body contour. By extracting the contour and the smallest external rectangular of the human body, the height-width ratio, perimeter ratio, duty ratio, can be calculated as human motion characteristics.In the field of human behavior recognition, human behavior are identified and grouped by adopting1-v-1SVM with different kernel function in this paper. At last, we compare the recognition rate of different kernel functions.