RANSAC直线估计方法在路面三维点云优化中的应用  

Application of line estimation algorithm based on RANSAC in road points cloud optimization

作　　者：曹毓[1] 冯莹[1] 杨云涛[1] 赵立双[1] 

机构地区：[1]国防科学技术大学光电科学与工程学院,湖南长沙410073

出　　处：《红外与激光工程》2012年第11期3108-3112,共5页Infrared and Laser Engineering

基　　金：军队预研项目(BY2008J018)

摘　　要：在机场跑道异物检测和公路路面病害勘测等应用领域,快速高效获得大面积路面三维形貌至关重要。在利用车载单线式激光扫描仪实施路面三维重建的应用中,车辆的颠簸会使得路面三维点云数据出现失真。为此,使用RANSAC直线估计算法估计出由扫描仪获得的每条扫描线的斜率,进而消除了路面三维点云的不规则起伏失真。实验结果表明,优化前由车辆颠簸导致的路面点云起伏幅度达0.6 m,优化后的路面点云起伏幅度控制在0.04 m以内。路面重建效果经优化后有较大改进,此外点云优化算法耗时约11 s。因此,文中的路面三维点云优化方法原理简单、实施简便、路面点云优化质量高且耗时少,可在车辆存在颠簸时保证路面三维重建效果,在需要快速高效获得大面积路面三维形貌的场合具有较好的应用前景。It is important to get top-quality three-dimensional shape of road surface fast in the applications of airport foreign bodies detection and road disease reconnaissance. In the application of road surface 3D reconstruction with vehicle-borne laser scanner, the balleting of moving vehicle can induce the distortion of the road surface points cloud. On this ground, the line-estimation algorithm based on RANSAC was used to get the slope coefficient of scan lines. After that, the anomalous fluctuation of road points cloud was removed. Experiment result shows that, the fluctuation amplitude of road points cloud before optimization induced by balleting of moving vehicle was 0.6 m. After optimization, the fluctuation amplitude of road points cloud was less than 0.04 m. The three-dimensional shape reconstruction effect of road surface was improved after arithmetic optimization. Besides, running the optimization arithmetic took 11 seconds. Therefore,the point cloud optimization method on road surface proposed in this paper has the advantage of simple principle, convenient using, high-speed calculation. The 3D reconstruction effect meets the reauirement in oractice_

关 键 词：点云优化 路面三维重建 RANSAC算法 直线估计 

分 类 号：TP212.9[自动化与计算机技术—检测技术与自动化装置]