Background extraction algorithm base on Partition Weighed Histogram

The common method to detect moving objects in image processing and Intelligent Transportation Systems (ITS) is background subtraction. So the background extraction is an important aspect of traffic vision detection and analysis. In order to improve the accuracy and the applicability of background extraction, a new background extraction algorithm is presented, which is called Partition Weighed Histogram (PWH) method. And in order to compare PWH method with other algorithms, an evaluation rule for comparison is built. Finally, different traffic videos are taken to verify the performance of PWH algorithm comparing with other algorithms.