基于改进Hough变换的结构化道路车道线识别

针对现有结构化道路车道线弯道检测识别技术的准确性和鲁棒性不高的问题，提出一种基于改进Hough变换的车道线识别方法。首先利用Canny边缘算子进行车道边缘检测，其次对比相邻区域距离内的Hough变换峰值参数值，将小于设定阈值距离的直线段进行连接，拟合形成车道线检测区域，然后根据消失点的横坐标距离图像中心点的位置来预测车道线方向，最后借助MATLAB平台完成车道线的识别。验证结果表明:该方法避免了曲线模型复杂、计算量大的缺点，实现了直、弯车道的识别统一化，识别准确率为95.3%，平均耗时0.036s，具有很好的实时性和鲁棒性。

Structured Road Lane Line Recognition Based on Improved Hough Transformation

Aiming at the problem of low accuracy and robustness of existing detection and recognition technology for lane bends on structured roads, a lane recognition method based on improved Hough transform is proposed. Firstly, Canny edge operator is used to detect lane edge. Secondly, the peak value of Hough transform parameter in the distance between adjacent areas is compared, and the straight line segments less than the set threshold distance are connected to form the lane line detection area. Then, the direction of lane line is predicted according to the distance between the abscissa of vanishing point and the center of image. Finally, the lane recognition is completed with MATLAB platform. The verification results show that, the method avoids the disadvantages of complex curve model and large calculation, realizes the unification of straight and curved lane recognition, the recognition accuracy is 95.3%, the average time is 0.036s, which has good real-time performance and robustness.