中文说明： 帧差法是最为常用的运动目标检测和分割方法之一，基本原理就是在图像序列相邻两帧或三帧间采用基于像素的时间差分通过闭值化来提取出图像中的运动区域。首先，将相邻帧图像对应像素值相减得到差分图像，然后对差分图像二值化，在环境亮度变化不大的情况下，如果对应像素值变化小于事先确定的阂值时，可以认为此处为背景像素:如果图像区域的像素值变化很大，
English Description：
Frame difference method is one of the most commonly used moving object detection and segmentation method, the basic principle is in image sequences of two adjacent frames or three frames using based on pixel difference by close value to extract moving regions in the image. First, adjacent frame image corresponding to a pixel value is obtained by subtracting the difference image, then the differential image binarization, under the environment brightness changes little, if corresponding pixel value change is less than the predetermined threshold value, can be considered here as the background pixel: if the image region of a pixel value changes greatly, it can be assumed that this is due to the induced by moving objects in the image, mark the area as foreground pixels using labeled pixel area that can be used to determine the position of the moving object in the image. Because the time interval between adjacent two frames is very short, the background model of the previous frame image as