基于圆形特征点的非线性相机标定方法

针对高精度标定板制作困难费用高的问题,提出了一种快速简易的相机标定方法。用数码相机对一张具有不同大小圆形特征点的标定表绕着光轴旋转拍摄几幅图像,利用亚像素边缘轮廓检测算法检测图像中的轮廓;采用最小二乘椭圆拟合方法得到亚像素椭圆中心坐标,用稳定的图像点与空间点的对应算法确定图像点与空间点的对应,计算过程中对所求参数进行了非线性优化。实验表明,重投影平均误差在0．2个像素以下,证明了该方法的可行性和较高的标定精度。

Nonlinear Camera Calibration Method Base on Circular Feature Points

It is difficult and expensive to manufacture high-precision calibration template. To solve this problem, a quick and easy camera calibration method was proposed. Firstly a calibration table, which had different sizes of circular feature points, was drawn. And then several images of the calibration table were shot by rotating the camera around the optical axis. A sub-pixel edge contour detection algorithm was used to extract the contours of the images. Then least-squares ellipse fitting method was used to get the subpixel ellipse center coordinates. A stable corresponding algorithm was used to determine the correspondence between the image points and the space points. Finally camera parameters were calculated based on Zhang＇s calibration algorithm, the camera parameters were optimized during this process. Experiments show that the mean re-projection error is; less than 0.2 pixels. So the calibration method is feasible and has high calibration accuracy.