Traffic Flow Parameter Detection Based on Epipolar Plane Image

Abstract

Traffic flow parameters detection is one of the most critical parts of the intelligent traffic management system in currently. With the continuous development of computer vision and image processing analysis technology constantly development, video detection technology in the traffic parameters testing is more and more attention. Video automatic detection technology utilization of computer vision, image processing and analysis, pattern recognition technology, compared with traditional detection technology induction loop detector, microwave radar detector and magnetometer detector, etc. The video automatic detection has many advantages, such as simple installation, convenient maintenance, not to harm the roadbed, low cost, high accuracy, and wide monitoring area. Firstly, the image sequence can be acquired through the video automatic detection equipment mounted over the road, and then two two-dimensional (2D) spatial-temporal images, a panoramic view image and an epipolar plane image, are formed for each lane in this article. By using the methods of constructing a general epipolar plane image model, to reflect the vehicle surface characteristics and movement characteristics of the driveway. Then put forward the texture Canny texture edge detection algorithm of image based on Gabor filter, according to the texture edge detection of three steps: normalization processing, filtering processing and edge texture detection, analyzes the edge texture detection algorithm of epipolar plane image, and provides normalized epipolar plane image, presented the specific calculation method and formula of Canny texture edge detection based on Gabor filter. Finally, according to the graphics after processing, examines traffic flow parameters detection for traffic volume, vehicle speeds and lane occupancy and analyzes summary of the whole algorithm.