| Intelligent vehicle technique is a pop research area of ITS(Intelligent Transportation System), and image processing based road/lane recognition is one of key technologies of vehicle assistant/autonomous driving.The main purpose of this paper is to improve the speed of image processing of lane/road recognition. An integrated method based on monocular vision is proposed to detect the road/lane marking in front of the car and acquire the position and direction of the lane, also a monocular vision hardware and software system is developed.First, Grid is used to mark off the region of interesting in initial image, and then all the pixels are eliminated except those on the intersection of the ROI grid lines, thus feature pixels of the lane edge in these intersections will be detected and can be used to generate some dynamic windows with dilatation algorithm. Then, images are processed in these dynamic windows to obtain lane edge feature. Slope and intercept of the lane can be gained by Hough transformation. At last, the information obtained will be compare with prior information in order to obtain correct lane location parameter.Also, a rapidly calibration method is adopted to calibrate the camera. A calibration plate is used to obtain the internal and external parameters of the camera, realize the transition between image coordinate and world coordinate. The experimental results showed that the calibration and installation scheme was feasible.At last, the proposed algorithm was validated by the videos captured from real traffic road scenes, and the experimental results showed that the processing speed of structured road lane was about 20ms/frame, and the methods were efficient, stable and accurate. |
PDF Full Text Request