Design And Implementation Of Video Anomaly Diagnostic System

In nowadays society,video monitoring system has been used in kinds of fields widely all around us.With the drastic increase in the use of video monitoring system,it's becoming more and more difficult for monitors to maintain and manage these equipment.In a network video monitoring system,the acquisition,compression,signal coding,transmission,decoding and other conventional video information processing process,they will often generate all kinds of video data damage,and result in declining in the monitoring results,which will seriously affecting the quality of video image.Therefore,an real-time video surveilance system,which can automatic detect the anomaly of video,is of very important significance.To improve detect efficiency and decrease the working hours for monitors,video anomaly diagnostic system is designed by myself.The paper analyzes of the different and common image abnormalities,including video signal loss,noise,brightness,difinition abnormality,color cast,image freeze and camera interference and so on and so forth,and judges and evaluates them in accordance with corresponding detection algorithm.Signal loss detection is based on gray histogram,noise detection is based on SNR,brightness detection is average gray method based on HSV color space,color cast detection is equivalent circular method based on CIELab color space,definiton detection is based on “edge point”,image freeze is base on difference method and camera interference is based on ORB keypoint detection and matching.The system has four functional modules,which cooperate together,including image acquisition module,image processing module,image abnormalities detection module and abnormalities record and display module.Currently,the video anomaly diagnostic system has been developed,and we use monitoring video to test the system.Testing results indicate that the video anomaly diagnostic system is able to not only provide real-time monitor of video quality,but also provide accurately detection in the abnormalities types of images.