基于车载激光点云数据的路面检测与重建

摘要

In this paper , we took vehicle-borne laser point cloud data as research object and proposed a new method for road detec -tion and reconstruction .We extracted the path of scanning car using file of GPS or the rotation angle of the laser scanner .And then we used the extracted path to detect the road edge .Then the abnormalities of the generated road edge were smoothed .We can reconstruct the road surface by obtaining the three -dimensional model of road surface with the accurate road edge .The three-dimensional model of road can be exported as 3D object file format to satisfy the demands of accurate road surface in related fields , such as three-dimen-sional city modeling .%以车载激光点云数据为研究对象，提出了一种新的路面检测与重建的方法。该方法根据GPS定位器中的索引文件或激光扫描仪的旋转角度值提取扫描车的车行路径，利用提取到的车行路径去检测对应的路沿。然后，对检测到的路沿中出现的各种异常进行相应的平滑处理。在得到精确的路沿线后，可得到路面的三维模型，从而完成对路面的重建。输出的路面三维模型为OBJ格式，可满足相关领域对精确路面的需求，比如三维城市建模。