Traffic Parameters Detection Using Edge and Texture

This paper proposes an approach towards the real-time detection of traffic parameters such as such as the stationary queue length and lane share in crossroad, based on the stationary camera installed along roadside. According to the characteristics of the traffic scene, Canny edge detection is used to get the edge information of region of interest (ROI). Local binary pattern (LBP) texture method is used to obtain the vehicle and road surface texture features and morphological processing is taken to enhance the responses of vehicle targets and reduce the noises generated by road lane and shadow interference. Then, edge and texture information is integrated to get the vehicle detection results, and inter-frame difference is taken to segment the moving and stationary vehicles. And then vehicle traffic parameters such as the stationary queue length and lane share are extracted. Through experiments taken in traffic command system of Dongguan City, the results show that this approach is accurate and performs well in real-time under different weather and illumination. (C) 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of Harbin University of Science and Technology