Study On Static Scene Surveillance Video Frame Removal And Moving Target Detection

In the past 20 years, video surveillance has undergone phase analog video surveillance,digital video surveillance and intelligent network video surveillance phase stage. The emergence of digital video surveillance, video surveillance for the quantitative analysis of computer data provides a great convenience.With the steady development of the video surveillance industry market, as well as favorable market expectations, high-definition monitor gradually replace SD monitoring,high-definition monitor has become an irreversible trend. However, the replacement of costly monitoring equipment, monitoring "massive" data, and the huge gap between theory and practical application of the resulting video, have become the further development of the video surveillance industry needs to face.Mass surveillance video data results in data difficult to effectively manage and analyze the huge amount of information contained in video surveillance, but little useful information.Normally, not all of the garbage can be executed and processing computer. However, the storage space is wasted static scene surveillance video frame memory repeatedly caused can be reduced. With the advent of the era of high-resolution monitor, the current complexity of the commonly used algorithms moving target would need to be reduced. In this paper, for static scene surveillance video frame removal and moving target detection two aspects were studied.In the surveillance video frame to remove static scene: the removal of surveillance video frame static scene, not only can reduce the amount of video data monitoring, cost savings, but also to shorten the discovery back to see surveillance video concerns the time necessary.Through the inter-frame difference threshold analysis, we found that the timeliness of the inter-frame difference method is high, be able to monitor the timely detection of the scene.Through the analysis of the characteristics of the frame difference detection pixels to construct a method for inhibiting the detection point, after such treatment can reduce the storage approximation of still frames. Through the single exception information frame analysis to construct a single exception handling mechanism frame can be further reduced without abnormal pieces of information. Tests showed that mentioned in this article surveillance video still frame removal method is feasible.Moving object detection: combined with the Gaussian law, advantages yards Law,inter-frame difference method and the statistics method several methods proposed an efficient two-platen moving target detection methods. Through the discussion of the method ofupdating process templates and dual-mode version of the update mechanism, and illustrating the principles of the dual-mode version of the moving object detection. Tests showed that the template mentioned herein double moving target detection method can not only effectively detect moving objects, and can effectively remove the monitor moving objects in the scene by stopping the movement of time is too long to be integrated into the background after the sudden re-generated by the movement "ghosting".