Research On The Examination Room Monitoring System Based On SIP Protocol And Video Summary Technology

As a common way to select talent,examinations are in a variety of ways.In order to ensure the impartiality and transparency of it,the examination room monitoring system came into being.Traditional examination room monitoring system is extremely inefficient.It is especially important to achieve a examination room monitoring video quickly browsing,get useful information in a timely manner,lock suspicious cheating objects in summarizing the traditional video.As an important protocol in the application layer,SIP protocol makes it possible to realize the mutual establishment or termination of the multimedia data,which could in a good compatibility and expansibility.Therefore,it is of great significance to apply SIP protocol and video summarization technology to the examination room monitoring system.In this paper,the main work and innovation are as follows.(1)The paper firstly analyze the SIP protocol,the RTP / RTCP protocol,and then use the SIP protocol to make the UA and proxy server keep in corresponding dialogue.And finally analyzed the SIP protocol session mode and process through the Wireshark and FreeSwitch.(2)Gets monitor video by the SIP camera,the paper introduces the technical system,implementation method,the main function of the system.Then the client can communicate with the server through the SIP protocol.Thus the various parts of the examination room monitoring system could be connected with high efficiency.Eventually it is the concept and technology of video summarization(3)This paper deals with the noise problem in the image,and introduces the background subtraction method,the frame difference method and the optical flow method for the moving target detection.In this paper,an improved method of Gaussian filtering is presented,in order to suppress noise,smoothing the image and making the image edge clearer.And for the background difference method in the "missed" situation,the paper proposed an improved hybrid Gaussian model which can improve the correctness of the detection.(4)Based on the moving target tracking algorithm Meanshift and Camshift,the RGB color space is converted to the HSI color space.Compared to the RGB color space,HSI color space is more intuitive on the perception of color,simultaneously solve the vulnerable to the impact of the surrounding environment defects.(5)A video digest system algorithm is designed.Using the C + + and the OpenCV open source library to combine other related library,the digest system algorithm could track and detect the activities of the large range of activities students.