一种利用点云强度特征的交通标志线提取优化算法

An optimal algorithm for extracting traffic lines using point cloud intensity features

针对车载道路点云中交通标线的提取问题,设计了一种联合强度特征图像和边缘检测的交通标线提取方法。该方法首先将道路点云路面转换成点云强度特征图像并进行局部二值化操作,然后利用改进的Canny算子进行边缘检测获得强度梯度图像,最后进行连通区域分析实现对交通标线的精确提取。通过5组具有不同标线类型数据的实验结果表明,该方法提取交通标线的正确度、完整度和精度均超过90%,证明了本文方法的有效性和适用性。

Aiming at the problem of the extraction of traffic markings in the vehicle-borne road point clouds,a traffic marking extraction method combining intensity feature images and edge detection is designed.This method first converts the road point cloud pavement into a point cloud intensity feature image and performs a local binarization operation,then uses the improved Canny operator to perform edge detection to obtain the intensity gradient image,and finally performs connected area analysis to realize the analysis of traffic marking precise extraction.The experimental results of five sets of data with different marking types show that the accuracy,completeness and precision of the method are all over 90%,which proves the effectiveness and applicability of the method in this paper.