基于控制场的高精度相机标定技术研究

Research on high-precision camera calibration technique based on control field

相机标定是高精度视觉测量的基础,是视觉测量的关键技术。相机参数的标定精度直接影响最终的测量精度。针对大视场的高精度测量,文章提出了一种基于控制场的相机标定方法。首先,通过构建相机参数求解模型,利用DLT(直接线性变换法)获得相机参数初值,再利用L-M最优化方法求解最优化模型参数即要标定的相机参数。本标定方法的标定结果平均反投影误差达到0.1个像素点,表明本标定算法具有很高的标定精度,可以满足视觉测量的高精度测量需求,在工业测量中具有较高的使用价值。

Camera calibration is the basis for high-precision vision measurement,and is the key technology of vision measurement.The accuracy of camera parameters directly affects the final measurement accuracy.For high-precision measurement of large field of view,this paper proposes a method for the camera calibration based on control field.First,by building a model of camera parameters to solve,use DLT(direct linear transformation)to obtain initial parameters of the camera,then use L-M optimization method for solving the optimization model parameter that is to be calibrated camera parameters.Results show that the average back-projection error can reach to 0.1 pixels by using the calibration method,and show that the calibration algorithm has high calibration accuracy to meet the needs of high-precision vision measurement,which has a high value in the industrial measurement.