作者: Qing Lin , Youngjoon Han , Hernsoo Hahn
关键词:
摘要: Lane detection can provide important information for safety driving. In this paper, a real time vision-based lane method is presented to find the position and type of lanes in each video frame. proposed method, hypothesis generated verified based on an effective combination lane-mark edge-link features. First, candidates are searched inside region interest (ROI). During searching process, extended edge-linking algorithm with directional edge-gap closing used produce more complete edge-links, features like edge orientation width select candidate pairs. For verification candidates, color checked enclosed by pairs YUV space. Additionally, continuity estimated employing Bayesian probability model length ratio. Finally, simple departure built detect departures locations image Experiment results show that work robustly real-time, achieve average speed 30~50ms per frame 180x120 size, correct rate over 92%.