Research on Detection Method of Abnormal Behavior of People in Video Surveillance

In view of the abnormal behaviors of all kinds of people, this paper starts with the types and characteristics of abnormal behaviors of people. After fully understanding the characteristics of abnormal behaviors of people, three groups of abnormal behaviors feature extraction methods are selected. The abnormal movement behavior, color and texture features of people in the monitoring are extracted, and on this basis, SVM is used to detect and identify abnormal behavior characteristics of people in video surveillance, in order to provide reference for the detection of abnormal behaviors of people in video surveillance.