基于机器视觉的车道标志线实时检测研究

Research of Real-time Machine Vision-based Detection of Lane Edge

本文研究了基于机器视觉的车道标志线实时检测问题。在讨论了各种车道模型的基础上,提出了一种基于分段切换的车道模型,并且简单论述了车道图像的检测算法,并进一步对试验效果进行了讨论与分析。试验结果表明,该基于分段切换车道模型的车道识别算法,能够很好地匹配各种形状道路,并具有较好的实时性。

Real-time machine vision-based detection of lane edge is studied. All kinds of lane models are described. A lane detec- tion algorithm based on the multi-sectional switch lane model is proposed and the detecting algorithm of lane image is discussed . The recognition results are given and analyzed in the end. Experimental results show that this method can fit lane boundaries with various shapes accurately and it is characterized by its good real-time etc.