In this paper, it is discussed how to detect the low SNR small moving target in the condition of complicated background. A space high pass filtering method is presented to improve the image quality for restraining the background noise and enhancing the small target, then the likelihood ratio detection theory is used to pick out a small number of possible moving targets and the neighborhood judging method is presented to pick out the smaller number of possible moving targets. Finally, the true target motion trajectory is detected in multi-frame successive image by use of image flow analysis method. The experiment result indicates that this algorithm can reliably detect small target's motion, even the point target's.