Research on Abnormal Behavior Extraction Method of Mobile Surveillance Video Based on Big Data

In the application of mobile surveillance video anomaly behavior extraction method, there is a problem of high unlocking rate. Therefore, a big data-based anomaly extraction method is designed. Segmenting moving surveillance video dynamic image, representing human body action in the form of mathematical symbols, extracting target feature key frames, matching two adjacent frames in video sequence, using big data technology to detect behavior trajectory, defining and distinguishing abnormal behavior, and improving abnormal behavior extraction process. Experimental results show that the average unlocking rate of the proposed method and the other two methods are 2.920%, 5.564% and 5.890% respectively, which shows that the proposed method is more effective.