Detection Of Abnormal Crowd Events Based On Video Analysis

In the modern society, the density of population becomes bigger and bigger and great accidents happen easily in crowded areas, which may lead to loss of life and personal injury. Lot of tragedies would be prevented if automatically monitoring of abnormal crowd incidents can be realized and those of which can be then reported to relevant personnel in time. Thus, there are significant meaning and prospect of applications in studying the abnormal crowd incidents.Analysis demonstrates that crowd behavior is very complicated, so in order to monitor it, the characteristics related to the crowd incidents must be extracted. Firstly, various studying methods of the estimation of crowd density are introduced and compared with each other. Then the definition of optical flow is presented and the basic theories of two algorithms, differential optical flow and block matching methods, are highlighted and the two are compared. According to the analysis of gang fights events and aggregation events, the crowd density of gang fights is big, movement is more violent and the direction of motion is haphazard while the crowd density of aggregation events is big, movement is gentle and moving target is less. Based on these differences, an algorithm of detecting the abnormal crowd incidents, which is used to distinguish gang fights and aggregation events from normal events, is advanced. The crowd density estimation algorithm based on the pixel counting is applied to the estimation of crowd density status in the video sequences so as to extract its characteristics. Then the differential optical flow method is introduced to obtain three characteristics that the estimation of crowd motion status, mean intensity of motion, variance of motion direction histogram and number of moving pixels. Lastly, analyzing of the testing video could provide the thresholds of these four characteristics, which can be used for the judgment of the video status.As the results show, the proposed algorithm, based on these four mentioned characteristics, can detect the gang fights and the aggregation events effectively. And the algorithm has a comparatively high recall and precision rate for detecting the gang fights and the aggregation events under complicated scenes.