Research And Implementation Of Video Surveillance System Based On Image Processing

Video surveillance system is widely used in various fields.It can be used not only in public such as residential area and prisons,but also in producing and managing systems such as industry enterprises,banks and electric power enterprises. In recent years,instable social events happen continually and people need more and more security.As an effective safeguard,video surveillance system has attached increasingly importance and stronger and broader application in every walks.Meanwhile,with the development of image digital technology,surveillance on important targets using image technology has become an indispensable part of our society.On the chance of the development of the image processing technology,this thesis studies a new thought and design of video surveillance system,background of the development of video surveillance system.In terms of hardware design,the USB camera is as an image capture,the EC5-1719CLDNA is as an control equipment,and the GSM module designed as an execute equipment.The function of the hardware design of every part is introduced in the thesis detailedly.In terms of software design,the thesis designs and implements a practical system consisting of several functional modules such as video image capture, video image pre-process,motive target detection and remote alarming,using Visual C++ and OpenCV.The moving target detection algorithm is the core of video surveillance system. Focusing on the comparation and improvement of every available algorithm,this thesis developes a new motion detection method,by combining the background subtraction and frame difference methods together with a adaptive background-updating method,in which several image frames are processed comprehensively.This method bears not only the characteristic "exact to survey object" of background subtraction but also the characteristic "robust to lighting variations" of frame difference.Further considering the adaptive background-updating arithmetic,the proposed method is supposed to be able to improve obviously the stability of a video surveillance system,especially one used at night.The system is steady with low wrong alarm percentage.All of these efforts have prepared to discuss farther problems such as motive target tracking.