New Algorithm for Extracting and Tracking of Video Moving Object Based of Spatio-temporal Information
一种时空联合的视频运动目标提取与跟踪新算法

A kind of new algorithm based on temporal and spatial union information is proposed.In spatial domain,an improved watershed is proposed to decrease the oversegmentation;in temporal domain,after compensating the global motion made by camera motion,this algorithm uses the accumulated sum of difference between a group of frames to reduce the segmentation error by using only two frame's difference,and adaptively selects the threshold of binary frame difference image.After the fusion of the result of temporal domain and spatial domain,the object can be extracted.Based on the sub-block of object region matching technique,a compact motion object tracking algorithm is proposed.Experimental results demonstrate that this algorithm can offer more accurate,faster and more effective extracting and tracking of video moving object.