PLANE FITTING OF POINT CLOUD FROM 3D LASER SCANNING BASED ON COOK DISTANCE
Yan Jianfeng 1,2) ; and Deng Kazhong 1,2)
1)Key Laboratory for Land Environment and Disaster Monitoring of SBSM, Xuzhou 221116;2)School of Environment Science and Spatial Informatics,CUMT, Xuzhou 221116
Abstract:Traditional leastsquares can not ensure high accuracy and stability for plane fitting when many outliers and error are mixed in original data. To solve this problem, a leastsquares method based on Cook distance is proposed. The method determines strong impact points by Cook distance value from point cloud data optimally and then accurately fitted plane can be achieved by leastsquares. The result of experiment and analysis shows that the fitted plane is closer to the real plane. The method can improve accuracy and stability of the process of point cloud reducing and model building effectively.