摘要
针对不同姿态下的三维可变形物体特征点匹配问题,提出了一种基于热核信号特征和测地距离的三点匹配策略,从而可以为后续的紧密匹配奠定基础.该方法主要由以下三步完成,首先提取出三维可变形物体的外部特征点;其次,对外部特征点采用热核信号定义局部特征;最后,结合局部特征描述符和特征点之间的测地距离进行最优匹配搜索,得到正确的匹配结果.实验结果表明:该方法能很好地实现对三维可变形物体的稀疏匹配.
This paper addresses the problem about the feature points matching of 3D deformable objects in different poses. It proposes a new algorithm called three-point matching strategy, which based on heat kernel signature （HKS） and geodesic distance. It includes three steps. Firstly, the algorithm extracts the external feature points from input 3D meshes. Secondly, the algorithm uses HKS to define the local feature for external feature points. Finally, the algorithm searches the optimal Geodesic distance matching between local feature descriptor and geodesic distance of external points. The experiment shows that the proposed algorithm can well realize the sparse matching of 3D deformable objects.
出处
《浙江工业大学学报》
CAS
2013年第5期539-544,共6页
Journal of Zhejiang University of Technology
基金
浙江省自然科学基金资助项目(Y1110780)
浙江省科技厅优先主题资助项目(2009C11038)
关键词
特征点提取
热核信号
测地距离
稀疏匹配
feature point extraction
heat kernel
geodesic distance
sparse matching
