首先采用灰度投影匹配进行全局运动估计和补偿，用以消除背景变化的影响；然后由二次差分抽取中间帧解决遮挡问题，通过Fisher 评价函数结合数学形态学填充得到运动对象分割掩膜，同时消除残余噪声以及平滑边缘。该方法在一定范围内较好地解决了遮挡问题，并能够高效快速地得到比较精确的视频对象。-first used for overall projection matching motion estimation and compensation, in order to remove the impact of changes in background; then by the second difference from the middle frame settlement block, Fisher evaluation function by combining mathematical morphology be filled Moving Object segmentation mask, while eliminating residual noise and edge smoothing. The method in a certain range to help solve the problem of shelter, and to the rapid and efficient access to more precise video objects.