摘要: 根据建筑物在高度方向截面上的点云数据必定位于其轮廓线的原理,提出基于聚类平面特征的点云数据精简算法。该算法无需对扫描对象进行表面重构,而是在保持建筑物高度方向数据精度的前提下,对点云数据分层聚类简化,保留满足条件的特征点,删除其余的点。通过实例证明该算法可以在保持建筑物外形特征的同时,达到较高的精简比率。
关键词:
三维点云,
聚类,
平面特征,
轮廓线,
数据精简
Abstract: According to the theory that the scattered point cloud data of buildings is certainly located on their contour line, this paper proposes a data reduction algorithm of clustering plane feature so that the scattered point cloud data of large group of ancient buildings can achieve a higher reduction ratio on the basis of maintaining the shape feature. Appling it to reconstruct the Small Wild Goose Pagoda with point cloud techniques achieves good modeling results.
Key words:
3D scattered points cloud,
clustering,
plane feature,
contour line,
data reduction
中图分类号:
王茹, 周明全, 邢毓华. 基于聚类平面特征的三维点云数据精简算法[J]. 计算机工程, 2011, 37(10): 249-251.
WANG Ru, ZHOU Meng-Quan, GENG Yu-Hua. Reduction Algorithm for 3D Scattered Points Cloud Data Based on Clustering Plane Feature[J]. Computer Engineering, 2011, 37(10): 249-251.