An improved method for motion detection by image difference

This paper makes improvements on motion detection method using image difference, specifically in: using the frame difference and background subtraction methods to detect motion. The cross-correlation method is used to detect slow movement and even the stopped target. Updating the background model to prevented the model from degradation. The experimental results show that the improved methods have significantly solved the problems of miss detection of slow movement and the background model degradation. Our method will lay a theoretical foundation for the behaviour understanding in moving target.