一种改进的点云边界快速提取方法

An Improved Method of Rapidly Extracting Boundary Points from Point Cloud

为进一步提高提取边界点的效率,提出了一种k-均值聚类与象限识别相结合的点云边界快速提取方法,通过k-均值聚类将点云划分为许多个子集群,根据三维格网划分方法探测出边界集群,在边界集群中通过象限识别提取出边界点。两组边界提取实验结果表明,本方法不仅能够提取到分布均匀的边界特征点,且与传统方法比较,效率更高。

Extracting boundaryfeatures from point cloud is an important work of data processing.In this paper,in order to accelerate the efficiency of extracting boundary points from point cloud,an improved method is proposed based on the combination of k-means clustering and quadrant recognition.Firstly,point cloud is divided into many sub-clusters using k-means clustering algorithm.Secondly,the boundary clusters are detected with the method of three-dimensional mesh division.Then,boundary points are recognized and extracted from the boundary clusters by means of quadrant recognition.Finally,two experiments are conducted to verify the feasibility of the proposed method.The results show that the proposed approach performs better than the traditional method in the aspects of extraction results and efficiency.