基于反演变换的双圆区分与拟合算法

Fitting Algorithm Based on Inversion Transformation for Data Synthesis of Double Circles

基于反演变换及完全最小二乘法拟合圆曲线方程,并依据数理统计方法,提出一种双圆的区分与拟合算法。该算法通过迭代实现,易操作而且效率高。数值例子表明,文章提出的算法能够准确地实现双圆的拟合与区分。

In this paper, we proposed an algorithm for distinction and fitting of double circles. The algorithm is based on inversion transformation and Total Least Squares （TLS） , and the hybrid point sets are separated in accordance with statistical theory. The algorithm is achieved iteratively with easy operation and high efficiency. Computation results show that the data set can be distinguished completely, and the accuracy of fitting results can be guaranteed.