The Study On Key Technologies In Intelligent Video Surveillance System And Their Realization

Intelligent video surveillance is a hot issue for the present study. It is added withmany intelligent module on the basis of traditional digital video surveillance. Thefront cameras collect the video images and send them to the system for furtherprocessing. People can get the information from the processed images. Whenabnormal situation arose, the surveillance system can identify the emergence of theunusual circumstances, and start exception handling system. Intelligent videosurveillance was widely concerned by many researchers for its broad range ofapplications. A variety of intelligent video surveillance software algorithms wasdeeply researched and tested. and directly promote the rapid development of relevanttheory.The key point is the smart video surveillance monitoring software algorithmdesign. When the video image was sent to the surveillance system, the computer canprocess a variety of video images and get the desired information. When abnormalsituation arose, the surveillance system can identify the emergence of the unusualcircumstances, and start exception handling system. All of these need the support ofthe fast and efficient computer software. The design of intelligent video surveillancealgorithms generally require digital image processing, target detection, facerecognition and other algorithms support.Digital image processing is the foundation of video and images software.Generally, the images got by the front cameras need the pre-processing of digitalimage processing and then then be processed by advanced algorithms for furtherprocessing. Digital image processing algorithms include image smoothing used forremoving the noise in the original images, sharpening algorithm used for enhancingthe image contrast, edge detection algorithm used for extracting the image edges andcontours, Hough transform algorithm used for getting image characteristic shape.The main purpose of the monitoring system was able to effective detection andidentification of the need to detect the target object, and this occupies a largeproportion of the role of video surveillance. Various methods of target detection include the background subtraction method and the optical flow method. Experimentsshow that the effect of these two methods in the detection of static images anddynamic image of the target has a good deal. These two algorithms is the basis of themore advanced video surveillance algorithm based system.Face recognition technology is the key technology of intelligent videosurveillance system. Face recognition is positioning the location of face, identifyingthe facial features of the person’s face and then identity recognition the owner of theface identity recognition from the images with faces. The face detection algorithmclassified a variety of ways. Haar features France is the main block to check the imagein the region through the establishment of boosting decision tree. Trained by manysamples, the system can be more precise identification of the face image. TheOpenCV containing the corresponding face image recognition library functions basedon Haar features method for detection processing.Intelligent video surveillance contains a lot of knowledge, and wide range ofareas, Therefore, the study of intelligent video surveillance algorithms is still going on.As technology continues to progress, I believe there will be more and moreconvenient and practical intelligent video surveillance products into the market, allowpeople to enjoy the charm of the high-tech.