一种三维激光扫描点云数据的预处理方法  

A pretreatment method of terrestrial 3D laser scanning point cloud data

作　　者：郝刚[1] 胡鑫凡[1] 岳朝林[1] 

机构地区：[1]黄河勘测规划设计有限公司,郑州450003

出　　处：《测绘科学》2014年第7期90-93,共4页Science of Surveying and Mapping

摘　　要：本文针对三维激光扫描数据存在异常数据和随机误差等特点,改进了两种点云数据的粗差剔除方法,针对点云数据的高频特性设计了基于卷积的等高线滤波方法,并通过实验证明其有效性;对扫描过程中数据可能存在缺失的情况给出了处理方法,并针对点云数据的质量做出了评定,讨论了利用强度信息进行数据滤波的可行性。According to the features of the terrestrial 3D laser scanning point cloud data as existing abnormal data and random data, this paper improved two gross error elimination methods of point clouds. Directing at the high frequency characteristics of point clouds, the paper designed the contour filtering method based on convolution and vestified the validity by experiment. Moreover, it put forward a method to deal with the loss data, and evaluated the precision of point clouds. Finally, it discussed the feasibility of applying the strength information in filtering.

关 键 词：三维激光扫描 点云数据 预处理 异常数据 滤波 

分 类 号：P232[天文地球—摄影测量与遥感] TP751[天文地球—测绘科学与技术]