A Novel Binocular Stereo Vision-Based Approach of Cycle Path Detection

In this paper, we describe a novel stereo vision-based approach of cycle lane detection. Bicycle lanes are important components of road scenes. Therefore, how to verdict the existence of bicycle lanes existence and to fix their positions is worthy of research. However, there is little research focusing on this topic. This approach combines lane lines detection with binocular stereo vision's advantage. The approach can judge whether there are bicycle ways or not, mark cycle lanes through cycle lane lines detection. Besides, for stereo vision has advantages at three-dimensional reconstruction, using binocular stereo vision can help us get cycle lane's position relative to cameras through binocular stereo vision more easily. Almost 200 sets of binocular pictures of different kinds of cycle lanes are collected to verify the feasibility of this approach. Experiments shows that this method can detect cycle lanes and gain its spatial location effectively.