Research On Moving Object Detectiona Nd Shadow Removal In Video Sequcnce

Moving Objects Detection based on video is a important component of the Video SurveillanceResearch. Intelligent Video Surveillance System distinguishes the moving information from thevideo sequence taken by camera; then detects, tracks and identifies the moving objects from thebackground, thus it can achieve the goal of determining the moving objects' behaviors. Therefore,the accuracy and timeliness of the detection of moving objects is essential for the whole follow-uptreatments of the Intelligent Video Surveillance System.This paper focuses on the following aspects.At first, the paper studies the Moving Objects Detection Algorithm based on backgroundsubtraction under the condition of fixed camera. This algorithm constructs the background frame bytaking the advantage of the frame difference's strength of effectively separating background inimage blocks, and presents the selective background updating based on scene dynamic change toadapt to the changes in time and space, and extracts the moving objects by means of backgroundsubtraction using adaptive threshold. Experimental results show that this algorithm can detectmoving objects effectively and quickly.Secondly, to deal with the problem that the result of detecting the moving objects is ofteninfluenced by their own shadow, Shadow Removal Algorithm in gray space is utilized in this thesis.First of all, this kind of algorithm applies the gray ratio to determine the shadow area preliminarily;and then it extracts shadow area precisely according to the cross-correlation and features of shadowratio edge; furthermore, it improves the Fast Normalized Cross-Correlation algorithm to reducecomputation.At last, this paper accomplishes the algorithm in the environment of VC++6.0 and VFW.Experimental results certify that the algorithm can effectively solve the problems of the influence ofbackground change, illumination change, and the moving shadow in an effective way.