一种时空联合的视频运动目标提取与跟踪新算法

New Algorithm for Extracting and Tracking of Video Moving Object Based of Spatio-temporal Information

提出了一种时空联合的视频运动目标提取与跟踪新算法。在空域分割中,针对分水岭方法过分割现象明显的缺点,对分水岭分割方法进行了改进;在时域分割中,首先对全局运动进行了补偿,随后为消除仅用两帧帧差进行对象分割所带来的误差,采用多帧帧差求和的方法,并自适应选取累积帧差的二值化阈值;时空分割结果进行投影融合后得到视频对象,提出用一种基于区域子块匹配的方法跟踪视频对象。实验结果表明,该算法简洁有效,能较好地把对象从运动背景中提取出来,并实现跟踪。

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 over-segmentation; 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.