Background extraction and update method based on histogram in YCbCr color space

Abstract

In computer vision-based Intelligent Transportation Systems (ITS), one of the key techniques is to exactly detect the vehicles. The common way for vehicle detection is the background difference method. In this paper, we introduce several background extraction and updating methods and develop a new background extraction and update method based on histogram in YCbCr color space. By using YCbCr color space, the influence of illumination change and shadows is reduced. To solve the problem with change in background itself, we propose a background update method by using the pixel change count and histogram. Experiment results show that the proposed algorithm can effectively extract and update the background information in complex traffic environment.