微细特征三维测量系统标定的精确椭圆提取方法

Accurate ellipse extraction method for calibration of micro-feature 3D measurement system

针对目前小微测量系统标定中圆心提取精度低与稳定性差的问题,提出了一种精确的椭圆中心提取算法。利用Zernike正交矩进行亚像素边缘检测,通过矩阵的块分解进行椭圆拟合。根据提取的亚像素边缘数据点,计算出最佳椭圆参数,提取中心。采用张正友的平面标定算法进行了标定实验,结果表明:所提算法精度高、稳定性好、易于实现。

Aiming at the problem of precision and stability of circle center extraction in the calibration of micro-feature three-dimensional measurement system,an accurate ellipse center extraction algorithm is proposed.Firstly,Zernike orthogonal moments are used for sub-pixel edge detection,and then ellipse fitting is performed by block decomposition of matrix.According to the sub-pixel edge data points extracted before,the best ellipse parameters are calculated and the center is extracted.The calibration experiment is carried out by using Zhang Zhengyou’s plane calibration algorithm.The experimental results show that the algorithm has high precision,good stability and is easy to implement.