基于仿射参数估计的迭代点匹配算法

本文提出了一种新的迭代点匹配算法。算法建立点集间仿射映射关系，把匹配问题转化为函数优化问题，通过点集间匹配对应关系和仿射变换参数的反复迭代最终求出问题的解。文中提出了构造虚拟点对和最小方差两种仿射参数估计方法，并利用改进最近点原则求解点集匹配关系，且证明了算法的收敛性。本文算法较好地解决了由仿射带来的非刚性形变点集匹配问题，且有很好的抗噪声和点性能。实验证明了算法的有效性和鲁棒性。

A Novel Iterative Point Matching Algorithm Based on Affine Parameter Estimation

A novel algorithm is proposed in this paper to solve the point matching problem.We suppose the point sets are in the relationship of affine transformation,and give out two affine parameters estimating methods,virtual pairs and least square methods.Based on this estimation utilizing the modified closest point rule,the algorithm iterates for the final exact affine parameters and corresponds.In the paper,the convergence of the algorithm is proved.Experi- ments on both synthetic point sets with varying degrees of affine deformation,noise and outliers,and on real image point-sets fully demonstrate the robustness of the algorithm.