发布时间：2010年7月4日 8时20分
摘要：针对传统数字化比对检测中点偏差计算的准确性及效率不高的问题，提出了一种基于约束搜索球的点云数据与计算机辅助设计（Computer Aided Design, CAD）模型比对检测技术。首先，为了提高计算的准确性，在分析现有点-面计算模型的基础上，提出了点-边计算模型及点-点计算模型，用于计算曲率变化较大区域的点偏差。其次，为了提高计算效率，在最邻近顶点搜索过程中采用K-D 树进行加速，在点数据归属过程中，构建一系列约束搜索球，将点云数据与CAD 模型比对的范围限制在约束搜索球内，降低算法的时间复杂度，提高点数据归属的效率。以某大型叶片为例，验证了该方法的有效性。
Precision Inspection of Point Cloud and CADModel Based on Constraint Search Sphere
Abstract: To improve the efficiency and accuracy of the existing inspection methods, a new inspection technology based on constraint search sphere was proposed. Firstly, Through analyzing the existing point-plane caculation model, the point-edge caculation model and the point-point caculation model were developed to inspect the regions with greater variation of curvature. Secondly, to raise the calculating speed, the K-D tree was used to search for the closest vertex on
Computer Aided Design (CAD) model, during the assignment of the point cloud, a series of constraint search spheres were constructed to restrict the scope of inspection. It decreases the complexity of the algorithm. The experimental results of a large leaf indicate that the method is accurate and efficient.
Key words: constraint search sphere; point cloud; computer aided design model; inspection
