摘要
针对三维碎片拼合中的碎片匹配问题,在提取碎片轮廓线的基础上,提出了一种快速的碎片匹配算法。该算法首先对轮廓特征点按其邻域曲面片的形状进行分类;其次根据特征点类型标志、曲率及特征段Hausdorff距离对不同轮廓线上的特征段进行相似性度量;最后利用法矢对相似度较高的轮廓段进行可匹配性验证。该算法较适合于用散乱点云表示的三维碎片模型。
On the basis of contour extraction of 3D fragments, an algorithm for rapid fragment matching in the 3D fragement reassembly was presented. In the algorithm, the type of feature point was calculated according to its neighbor surface type, and then higher similarity contour segments belonging to different contours were found, thus the normal vector was utilized to verify the matching probability of similar space curves. The similarity between feature segments of different contours was measured according to feature point type, feature point curvature and the hausdorff distance of feature segments. This algorithm is suitable for point-based surface.
出处
《计算机应用》
CSCD
北大核心
2008年第2期473-476,共4页
journal of Computer Applications
基金
国家自然科学基金资助项目(60475021)
河南省杰出青年基金资助项目(0412000400)
关键词
碎片拼合
碎片匹配
曲率
特征点
fragement reassembly
fragment matching
curvature
feature point