标定靶面平行成像平面时Tsai算法的改进

摘要

Aimed at the condition that the calibration board nearly parallels to the imaging plane, an improved algorithm ofTsai' s camera calibration is suggested. At first, a camera calibration model is built. Then, with the invariance direction of RAC, proportionality of RAC and principle of pinhole imaging, camera parameters are estimated by Levenber-Marquardt. At last, experiments are performed to verify the proposed algorithm. The experimental results show that the approach had a small amount of calculation, and can calibrate the camera parameters with good precision.%针对机器视觉应用中出现的标定靶面平行于摄像机成像平面的情况,提出对基于RAC(radial alignment constraint)摄像机标定算法的改进.建立了该种情形下摄像机标定模型,根据RAC约束的方向不变性、RAC约束的等比例性等特性,使用Levenber-Marquardt算法计算出部分参数,以及利用针孔成像原理求出余下的参数.通过实验对算法进行了验证和分析,结果表明,该算法计算量小,并且具有较高的标定精度.