基于几何特征的点云分割算法研究进展

点云是3维图像的一种特殊数据形式, 正逐渐成为3维图像信息处理的研究热点; 点云分割是点云数据处理的重要步骤, 对算法的结果有直接影响; 基于3维图像几何特征的点云分割算法结构简洁、运算结果稳定性强, 且易于调整, 在实际应用中占有主要地位。对最近几年涌现出来的基于几何特征的点云分割方法进行了梳理, 根据每种方法的理论基础和应用特点将算法归纳为基于边缘检测、表面特征和模型拟合的点云分割方法, 分析了各类算法的特点和存在的主要问题, 并进行了算法性能比较, 分析了影响点云分割算法效率的主要因素, 最后对未来的发展趋势进行了展望。

Research progress of point clouds segmentation algorithms based on geometric features

Point cloud is a special data form for 3-D image, which is gradually becoming a research hotspot of 3-D image information processing. Point cloud segmentation plays an important role in point cloud processing and has a direct impact on the results of the algorithm. Point clouds segmentation algorithm that based on geometric features of 3-D images are simpler in structure, more stable in operation results, and easy to adjust, which occupy a major position in practical applications. In this work, the point clouds segmentation methods based on geometric features emerged in recent years were sorted out. According to the theoretical basis and application characteristics of each method, the algorithms were classified into three categories: Edge detection based, surface features based, and model fitting based methods. The characteristics, the main problems of different algorithms, and the main factors that affect the efficiency have been analyzed. Finally, the algorithms performance have been compared, and the future development trend is prospected.