本文介绍了几种点云处理算法，包括基于区域增长的增量聚类、层次聚类、迭代简化及粒子模拟等方法。增量聚类通过不断添加最近的点到种子点直到集群达到最大尺寸；层次聚类采用自上而下的二进制空间分割方法；迭代简化和粒子模拟则是进一步简化和模拟点云数据的有效手段。
1,incremental clustering
#Clustering by growing regions
##start with a random seed point
##successively add nearest points to cluster until cluster reaches desired maximum size
2,hierarchical clustering
#top-down approach using binary space partition
#recursively split the point cloud if:
##size is larger than a user-specified threshold or
##surface variation is above maximum threshold
#split plane defined by centroid and axis of greatest variation
#replace clusters by centroid
1，
2，
3，
4，
5，
3,iterative simplification
4,particle simulation
扫一扫
1385
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
