Extract Semantic Context Information for Intelligent Video Surveillance of Traffic scenes

Abstract: Surveillance are nothing but an monitoring, here we use an video surveillance to provide Security to the people to avoid the unusual or abnormal activities in the field. Visual surveillance systems are mainly used for public security. we try to Extract here semantic context information which including object-specific context and scene- specific context information to make an intelligent system with robust object detection, tracking, warning signals, classification and abnormal event detection. In object-specific context information, objects are classified into an meaningful categories. Object classification is also called as a co trained classifier, which takes advantage of the multi view information of objects and collects the strong features to identify an object and reduces the number of labeling training samples, is learned to classify objects into pedestrians or vehicles with high object classification performance. For each kind of object, we learn its corresponding semantic features includes: motion pattern each vehicles, width distribution, paths, and entry/exist points. Based on this information, it is efficient to improve object detection, tracking, abnormal event detection and warning signals recognition.

Keywords: object detection, object classification, Gaussian mixture model (GMM) and graph cut, object classification, object tracking, warning signals, video surveillance.