基于改进坐标增量的点云数据压缩算法

西安科技大学 测绘科学与技术学院,陕西 西安710054

坐标增量法; 三维激光扫描; 点云压缩; 扫描线; 特征保留

Point cloud data compression algorithm based on improved coordinate increment
YAO Wan-qiang,LIN Xiao-hu,MA Fei,XUE Bei

(1.College of Geomatics,Xi'an University of Science and Technology,Xi'an 710054,China)

coordinate increment method; terrestrial laser scanning; point cloud compression; scan line; feature retain

DOI: 10.13800/j.cnki.xakjdxxb.2016.0615

备注

为提高点云数据三维建模及其应用的效率,在保证几何特征不变的前提下,进行数据压缩显得必要而迫切。针对地面三维激光扫描获得的点云数据密度大、冗余信息多,现有压缩算法存在不足的问题,在分析研究现有算法的基础上,将坐标增量法中一维扫描线点云数据逐点压缩扩展到二维扫描线与扫描线间点云数据的压缩,提出了改进坐标增量的点云数据精简压缩算法。并通过实例,借助Matlab平台编程,将该算法的压缩效果与坐标增量法、随机采样法、区域重心法和曲率采样法等现有典型算法的压缩效果进行定性和定量比较发现,对于按行或按列扫描的平面或曲面点云数据,该算法所用的时间较短,速度适中,且能很好的保留特征信息,具有较好的精简压缩效果,为大数据时代下海量点云数据的存储与管理提供了一定的参考。

In order to improve the efficiency of 3D modeling and application of point cloud data,it is necessary and urgent to carry out data compression under the premise of ensuring the geometric feature.The terrestrial laser scanning often produces high density and information redundancy of point cloud data.However,the existed algorithms are insufficient.Extending the point cloud data point by point compression of one dimensional scanning line in the coordinate increment method to two-dimensional scanning lines between,thus a point cloud data compression algorithm based on improved coordinate increment has been proposed on the basis of studying the existing algorithms.A case study is conducted in Matlab to compare the compression effect of the proposed method and several existing typical compressionmethod from qualitative and quantitative such as coordinate increment algorithm,random sampling algorithm,barycenter of area data compressing method and curvature sampling algorithm.The experiments show that the proposed method achieves good compression effect with a relatively short time,moderate speed and well preserved feature information for point cloud data of scan-lined plane or curved surface,providing some reference for the storage and management of massive point cloud data in the era of big data.