A lane recognition system based on priority

Using OpenCV, a lane recognition system is designed. In this study, a lightweight approach is proposed to real-time detection of lanes using ROI (region of interest) based on priority setting. The traditional ROI is further divided into two regions of different priorities. The left line of the lane, detected in the left region, is given a higher priority. And it can deduce the other line of the Lane using camera calibration. It implements the fast lane detection and generation combined with Hough transform. Experimental results show the method has significant high output effectiveness. It reduces the time of image processing, and improves the real-time performance of the system. Even if the road information completely lost, the system could mark lane for temporary help.