天津工业大学学报2025，Vol.44Issue(2)：72-77,6.DOI:10.3969/j.issn.1671-024x.2025.02.010
基于随机采样一致算法的卡车车厢点云分割
Point cloud segmentation of truck compartment based on RANSAC algorithm
摘要
Abstract
To address the challenge of segmenting truck bodies due to their diverse types and the difficulty in segmenting deformed compartments,a segmentation approach based on RANSAC is proposed.Initially,the compartment surface is evenly divided to obtain multiple planar patches.A Kd-tree is employed to establish topological and geometric relationships among these patch point clouds.The Euclidean distance from a point to a plane is used as the criterion for initial segmentation,with a distance threshold applied.Subsequently,an angle threshold is de-signed in conjunction with the RANSAC algorithm to optimize and merge the patches,achieving precise segmen-tation of various types of truck bodies.The experimental results demonstrate that the truck segmentation model developed can segment multiple types of truck bodies.Specifically,the maximum relative errors in measuring the dimensions of double-sided,four-sided,and five-sided trucks are 0.048,0.031,and 0.046 m,respectively.The measurement accuracy meets engineering requirements.关键词
物料装车/激光雷达/点云分割/RANSAC算法Key words
material loading/LiDAR(light detection and ranging)/point cloud segmentation/RANSAC algorithm分类
计算机与自动化引用本文复制引用
耿磊,杨梅,肖志涛,张芳..基于随机采样一致算法的卡车车厢点云分割[J].天津工业大学学报,2025,44(2):72-77,6.基金项目
天津市自然科学基金项目(24JCYBJC00310) （24JCYBJC00310）
