Motion detection is of paramount importance in video surveillance systems. In this paper, a novel algorithm is proposed to extract the exact boundaries of moving objects in video frames. Using the concepts of Cross-Correlation and Edge Detection, we combine two well-known motion detection methods to extract the moving regions more accurately. Also, we modify these two methods to in terms of accuracy and processing time in order to make them suitable for algorithm. The experimental results prove that the proposed method significantly outperforms the other state-of-the-art schemes in terms of both objective and subjective performance. Another advantage of the suggested method over the others is its relatively low computational complexity and high implementation speed.
Motion Detection, Background Subtraction, Moving Object Extraction, Video Processing