Abnormal crowd behavior detection based on motion effect map features of moving foregrounds

Abnormal behavior detection is one of the hot areas of research in intelligent surveillance field. An abnormal crowd behavior detection algorithm based on motion effect map features of moving foregrounds was proposed. Firstly, moving foreground segmentation algorithm based on adaptive GMM model was used to extract the foregrounds of the video sequence. Then each video frame was divided into blocks in order to achieve motion effect map of moving foreground blocks by acquired foreground area, and motion effect map features of each block could be achieved. An improved K-means algorithm by optimizing initial clustering centers was employed to train and test dataset. The experimental result shows that the proposed method effectively improves the accuracy of detecting unusual behavior in abnormal frames compared to existing algorithms, and the location of abnormal behavior can be located.