Research And Application Of Shape Context In Intelligent Video Surveillance

The video analysis on human and face is one of the hot and difficult computer vision technologies. The majority of monitoring objectives in video surveillance system are human. So the video analysis on human and human face is the foundation for video surveillance system to be intelligent.In this paper, we explore and research intelligent video surveillance system with the relevant technology, focus on the object recognition and face matching in intelligent video surveillance system. Shape matching is one of the basic issues in computer vision and pattern recognition. It can be used in many areas, such as object recognition, content-based image retrieval, etc..Traditional Shape Context algorithm is used for shape matching with simple shape. In this paper, we research the shape matching on the low vision of face in the application of intelligent video surveillance system, and apply Shape Context shape matching technology to face shape matching.In this paper, the contour extraction algorithm has been improved to meet the complexity need of the face contour. We use diffusion filter as preprocessing stage: help to remove the inside clutter edge of human body, make the contour more salient. Then we use the improved edge detection and contour extraction algorithm to enhance the important feature points extraction, and sample the feature points both less and better.The method is based on Shape Context algorithm, uses the two-dimensional invariance of log-polar transformation, and overcomes the shortcoming of the traditional shape matching algorithm. It is proved that this shape matching method is stabile and effective, and it is applicable to intelligent video surveillance system.Finally, Shape Context Face shape matching algorithm and other advanced image processing technologies are integrated into Chengfeng company's intelligent IP video surveillance system. Face Detection, Tracking and Alarm module can detect face, track and record their information. After face detection and recognition, the high-speed PTZ camera can track the suspicious, high-definition featured shoot, face shape match offline using Shape Context.