点到平面距离的点云数据压缩方法

A point cloud data compression based on the point to plane distance

针对三维激光扫描点云数据采集过密、冗余信息较多,现有压缩算法存在不足的问题,该文提出了基于点到平面距离的散乱点云压缩算法。将该算法与基于三角形法向量夹角和格网法两种现有算法的压缩结果进行比较,通过对比构建的空间三角网可以发现,该文算法对物体特征复杂的部位有较好的压缩效果,且在压缩率较高时,不会使较平缓的部位出现过度压缩而失真的情况。

The 3 dimension laser scanning technology is a new data acquisition method. For the dense data processing of point cloud, the paper presented a data compression method of scattered point cloud da- ta based on the point-to-plane distance. The new method of the compression effect was more ideal. The tri- angle net was reconstructed using compressed cloud data. Through comparing the triangle net with two typical algorithms： the adjacent triangle normal vectors included angle and the grid, it was found that this algorithm got preferable results in some substance with complex characteristic. Besides, when there was a higher compression ratio, it also helped to avoid the occurrence of distortion caused by excessive compression of gentle place.