结合超体素几何和色彩信息的快速3D点云分割

Fast 3D point cloud segmentation combined with super voxel geometry and color information

针对当前最先进的算法也存在因物体之间堆叠和遮挡关系造成场景分割速度慢、精准度低等问题,提出了一种新颖的快速3D点云分割方法,从点云的超体素划分开始,利用一种捕获几何和颜色信息的新度量标准合并超体素,以获得维持分割区域层次结构的3D分割;该方法还具有与输入点云大小呈线性相关的时间复杂度。与凹凸性方法和基于边界的方法相比,所提出的算法的分割精确度和效率更好。

To solve the problem of low segmentation speed and precision inducing by the stacking and blocking relationship between objects,a novel fast 3D point cloud segmentation method is proposed,which starts from the super voxel division of point cloud,and uses a new metric to capture geometric and color information to combine super voxels to obtain 3D segmentation that maintains the hierarchical structure of the partition.This algorithm also has a time complexity that is linearly related to the input point cloud size.Compared with the convexity method and the boundary-based method,the proposed method performs well in segmentation accuracy and efficiency.