非一致性稀疏采样的LiDAR点云压缩方法  

LiDAR point cloud compression method based on non-uniform sparse sampling

作　　者：陈元相 陈建[1] 郑明魁[1] 陈志峰[1] CHEN Yuanxiang;CHEN Jian;ZHENG Mingkui;CHEN Zhifeng(College of Physics and Information Engineering,Fuzhou University,Fuzhou,Fujian 350108,China)

机构地区：[1]福州大学物理与信息工程学院,福建福州350108

出　　处：《福州大学学报（自然科学版）》2021年第3期329-335,共7页Journal of Fuzhou University(Natural Science Edition)

基　　金：国家自然科学基金资助项目(62001117);产学研联合开发资助项目(20195GYCJS001)。

摘　　要：针对自动驾驶车载LiDAR点云,提出一种基于形态学分割和非一致性稀疏采样的点云有损压缩框架.将LiDAR点云分割为地面和非地面点云两部分,对两者进行不同强度的去冗余稀疏采样;然后将3D数据转换为2D形式的距离图像,并结合占据图和Morton排序,将点云表示为更加紧凑的1维距离向量形式;最后利用图像编码方法进一步压缩点云.实验结果表明,本文方法压缩性能明显优于现有算法,可达到更高的重建质量.Aiming at the point cloud of autonomous driving vehicle LiDAR,this paper proposes a point cloud lossy compression framework based on morphological segmentation and non-uniform sparse sampling.Divide the LiDAR point cloud into two parts,ground and non-ground point clouds,perform de-redundant sparse sampling of the two with different intensities.Then convert the 3D data into a 2D distance image,and combine the occupancy map and Morton sorting to sort the points.The cloud is expressed as a more compact 1-dimensional distance vector form.Finally,the image coding method is used to further compress the point cloud.Experimental results show that the compression performance of the method in this paper is significantly better than the existing algorithms,and can achieve higher reconstruction quality.

关 键 词：点云压缩 点云分割 球坐标变换 稀疏采样 Morton码 图像编码 

分 类 号：TN958.98[电子电信—信号与信息处理]