针对对称差分法检测目标时容易产生空洞以及当目标运动速度较慢或尺寸较小时易出现漏检等现象，提出了一种基于运动信息和标记多尺度分水岭的运动目标检测算法。首先将用高频强调滤波等处理的视频图像进行差分，再运用高阶统计运动检测算法检测出差分图像中运动目标的大概运动区域，经后处理得到运动目标的初始二值掩膜；并应用初始二值掩膜得到用于标记的前景与背景标识，用该标识修正当前帧的多尺度形态学梯度图像；最后进行分水岭分割，得到具有精确边界的运动目标。实验结果表明该方法能对运动目标进行准确检测，继承了变化检测和分水岭算法速度快的特点，克服了分水岭算法易产生过分割的缺点，且具有良好的鲁棒性。

The symmetrical differencing easily generates empty when checking objects, especially, it easily appears leaking of check when the object running slower or objects size is smaller, Hence, a method of motion detection based on motion information and marker multi-measure watershed is proposed. First, video image which processed by high frequency emphasis filter are treated using frame differencing.Second,the HOS(higher order statistics) is used for the difference image motion detection,then through post-processing,a initial motion object binary model can be obtained.Third,marker of foreground and background are derived from temporal information,which is used to modify multi-measure morphological gradient image.Lastly,watershed segmentation is used to acquire motion object with precise boundary. Experimental results show that the detection algorithm can detect object accurately.It has the virtue of low complexity inherited from change detection and watershed,and overcomes the detect that it is easy to engender the excessive segmentations of the watershed algorithm,so it has a high adaptability.