摘要
提出了一种特征保持的三维点云迭代简化算法。首先对点云模型构造KD树结构,计算采样点的k邻域,然后利用点云模型的局部几何信息作为参数,包括局部采样密度、采样点的精度和曲率,计算评估函数值,迭代删除评估函数值最小的点。实验结果表明,算法在简化点云数据的同时,能有效去除噪声数据,而且很好地保留了原始模型的特征信息。
This paper presented a feature-preserving iterative simplification algorithm for 3D point cloud model. It began with ~eonstrueting KD tree of the model, computed the k-nearest neighbor of sampling point. Defined local geometry information as the feature parameter, which considerd the local sampling density, sampling point accuracy, curvature. Then it computed the evaluation function value according to its feature parameter, iteratively reduced the point with smallest evaluation value. Ex- perimental results show that this method is effective to reduce the point cloud, remove the noise point and preserve the feature of the original point cloud model.
出处
《计算机应用研究》
CSCD
北大核心
2014年第4期1273-1275,共3页
Application Research of Computers
基金
国家科技支撑计划资助项目(2009BAI81B00)
关键词
点云简化
曲率
局部采样密度
评估函数
特征保持
point cloud simplification
curvature
local sampling density
evaluation function
feature-preserving
