Abstract: Space moving targets are often mistakenly detected even when they do not appear because the detection process is easily interfered by light, clouds and other factors. In order to solve this problem, we propose an improved frame difference algorithm, which combines the frame difference method with the background subtraction method and updates the current frame as the background frame periodically, namely, it makes the binarization processing to the images to be detected at the first step. This process can effectively eliminate such noise factors as light and clouds, and intensify the images of space targets, and then help the frame difference algorithm to detect space targets. According to the experiments, this method can effectively decrease the false detection rate of space moving targets. There is no need for the improved algorithm to put each updated frame as a background or make a statistical model of the background, so it improves the operating speed and simplifies the process of establishing the background. This algorithm is easy to implement and simple to operate. With low investment and high sensitivity it has great practical value.