中文说明：首先通过建立一个可靠的背景更新模型，然后在视频序列中取一帧图像与背景做差得到完整可靠的运动目标图像
English Description：
Firstly, a reliable background updating model is established, and then a complete and reliable moving target image is obtained by subtracting a frame image from the background in the video sequence