数字图像相关方法中的标定对三维形貌和变形测量的影响

Effect of calibration on 3D shape and deformation measurement in digital image correlation

为了实现三维数字图像相关方法的高精度三维测量,首先需要标定成像系统的内外参数.从数字图像相关的基本原理出发,针对相机成像的针孔模型,在理论上给出了物体成像的误差来源和修正措施,从而引出相机的各个标定参数.实验研究了多组标定板的姿态,分析了棋盘标定方法的稳定性和精确度,给出了标定过程的优化方案,解决了复杂工程环境下现场标定步骤繁琐的问题.实验结果表明:当标定板的姿态数大于9组时,标定参数趋于稳定,用稳定的标定结果计算目标物的半径更可靠.精确稳定的棋盘标定方法有利于获得高精度的三维测量结果,因此优化标定后的三维数字图像相关方法不仅操作简便,还具有相当高的可靠性和精确度.

To acquire high precision in three dimensional （3D） measurement, calibration of interior and exterior parameters for the imaging system is essential. Starting from the basic principle of digital image correlation （DIC）, the error sources and corrective methods of object imaging are given in theory for the pinhole model of camera imaging. In the experiments, multiple sets of chessboard calibration are studied based on the analysis of the stability and reliability of chessboard calibration. An optimized scheme for the calibration process is given. The process simplifies the complicated calibration steps in the field measurement. The experimental results show that stable calibration parameters can be obtained when there are more than 9 groups of calibration plate attitudes, which leads to more reliable results in the radius calculation of the experimental target. Therefore, an accurate and stable chessboard calibration method makes good contribution to high precision 3D measurement. Thus, 3D DIC with optimized calibration is proved to be an easy method with high reliability and accuracy.