Research And Application Of Intelligent Video Surveillance System

With the rapid development of the computer vision processing technology, electronic technology, communication technology, intelligent video surveillance system as an important means of defense and security is increasingly being attention. Intelligent video surveillance technology, including image processing, image analysis, pattern recognition, artificial intelligence and many other research fields .The process of intelligent analysis of visual surveillance consists of four basic steps: moving target detection, classification, tracking and video analysis, where moving target detection is the foundation. Moving target detection is to separate moving object from the background model .The quality of detection results, for objects classification, objects tracking, behavior understanding and other advanced post-processing plays a vital role. However, due to the dynamic changes of the background image, such as light, weather, shadows, and background interference effects, making moving target detection become a very difficult task. Therefore, Moving target detection is an important research focus in intelligent video surveillance under complex scene, its research has important theoretical significance and practical value.The main contents of this paper can be divided into three parts: Research and analysis of typical moving target detection algorithm; Proposed a new moving target detection algorithm and experimental simulation of the new algorithm; The applications of intelligent surveillance system in the substation.1. First, the paper introduces some basic moving o target detection method: optical-flow method, the time difference and background difference method, Second, the paper detailed analysis and comparison of a variety of methods, and point out its advantages and disadvantages and their respective scope of application.2. As the light slowly changes,the regional large-scale changes, object into or out of causes of the change of the background, In this case, the basic background difference method can not accurately detect the object. The paper proposes a responsible,efficient background model and update scheme,First, the algorithm use an improved background modeling based on Pixel gray-scale classification to initialize the background model. Second, the algorithm use a two-tier model adaptive background updating method based on background difference method for moving target detection, At the same time presents a simple shadow detection method to eliminate the shadow and dealing with the detected motion regions in order to obtain more accurate objects, And at the same time has obvious effect on the elimination of shadows. The algorithm is fast, accurate and has a wide range of adaptability; Finally, the paper used for the new algorithm, designed a set of Moving target detection system, the composition of the detection system modules and processes are introduced.3. On the basis of the research on the development process and requirement analysis of substation monitoring system.Paper design a set of substation intelligent video surveillance system.Program introduced in detail the overall framework of the system design, hardware component, software functionality and system characteristics. the implementation of the program can not only greatly improve power system security and efficiency but also have some reference on security systems of other sectors.