摘要
目前逆向工程中关于三维数据的平滑整定一般都是建立在三维数据之间的拓扑关系上的.提出一种平滑整定新算法,该算法是针对散乱、无序的,没有任何拓扑关系的三维离散点云数据.根据概率统计学中的频数法从大量的离散三维数据点云中找到噪声点,然后基于三维数据点云所建立的K-D树空间数据结构,找到噪声点周围的k个最近点,根据噪声点周边k个最近点的信息对噪声点处的真实信息进行恢复.基于激光三维平面扫描机器人系统证明该算法对离散三维数据点云的平滑整定的效果是令人满意的,在逆向工程中对数据预处理是切实可行的.
Currently, in reverse engineering, the data processing approaches for 3D point cloud data are based on data topology. The algrithm here aimed at the unordered, with no topology relationship, 3D scattered point cloud data. First, adopting the frequency method in probability statictics; the noise data could be found from the scattered data; then, through the K-D tree space estblished based on the 3D data, the k nearest points around noise point were found and through the information of these k points, the real information of the noise point was recovered. The laser 3D plane scanning robot system has demonstrated its effectiveness and it is feasible in pre-processing of data in reverse engineering.
出处
《工程设计学报》
CSCD
北大核心
2008年第2期128-133,共6页
Chinese Journal of Engineering Design
基金
国家自然科学基金资助项目(50575029)
