Vision-based Density Detection of Traffic Flow in Urban Intersection in the Daytime

In urban intersection of intelligent control system, detection of rows of vehicles is essential to the control algorithm, and density estimation method and vision-based detection algorithm are presented in this paper for sake of the limitation of using loop detectors method. Firstly, background segmentation and rebuilding is introduced and algorithm for vehicles segmentation in HSV color space is discussed. Then length and density of traffic flow are calculated. Finally, experimental results and analysis are included