Font Size: a A A

The Research Of Object Removal Detection Method In Intelligent Video Surveillance

Posted on:2009-07-30Degree:MasterType:Thesis
Country:ChinaCandidate:M M QiFull Text:PDF
GTID:2178360245470031Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
Video surveillance system is indispensable for the police, the traffic, the library, the museum, and the community security, etc. It can monitor, control, and manage the object intelligently, therefore it is one integrated manage system with the good ability of surveillance in practice. The intelligent video surveillance system can detect the motion of object in the specified area based on video sequences processing, and start alarm quickly when abnormal behavior is detected. The motion detection and alarm of the intelligent video sequence can replace the part of job of monitor personnel automatically and enhance the automatic level.Background Modeling is the basic way and object detection is the key important aspect of intelligent video surveillance system. There are five parts in the paper. The first part is a summary of the techniques of multimedia and surveillance system, which introduces their evolvement and relative technology. In the second part, we discuss some important techniques of how to design surveillance system, such as background model, background update, object detection. The research of object removal detection method in intelligent video surveillance system is discussed in the third part. In this paper, an effective object detection method is proposed, which uses Gaussian Background Model, combines area and gray-space histogram of object as its feature. This method can be used as object removal detection in library and museum, at the same time, the system can give sound alarm. After object removal detected, the system can learn new Background Model, so the intelligent video monitoring system can work on. The object removal detection method is implemented based on VC++6.0 with OPENCV, and tested with 150 frames. In the fourth part, it analyses the result of the detection method, points out its shortcomings and suggests possible improvements. In the end, the system and software are concluded and some technologies for the coming development are presented.
Keywords/Search Tags:intelligent video surveillance, background modeling, background update, object removal detection
PDF Full Text Request
Related items