基于一维标定物的高精度标定  

The high precision calibration of machine vision based on one-dimensional object

作　　者：王涛[1] 吕乃光[1] 杨剑[2] 

机构地区：[1]北京信息科技大学光电信息与通信工程学院,北京100192 [2]北京邮电大学电子工程学院,北京100876

出　　处：《北京信息科技大学学报（自然科学版）》2010年第1期21-25,共5页Journal of Beijing Information Science and Technology University

基　　金：国家自然科学基金项目(50675015);人才强教项目(PXM2007-014224-044674);重点实验室开放课题(KF20091123208)

摘　　要：基于一维标定物相机标定方法的标定精度问题,系统地分析了标定过程中的误差形成因素,提高精度的关键在于对畸变的校正。结合一维标定物的特点,提出用交比不变法对成像畸变进行校正。同时指出了在实际应用中应注意的一些问题,以提高算法的稳定性。实验表明,其标定精度可以达到基于平面标定法的标定精度,从而验证了算法的有效性。Based on the accuracy of camera calibration based on one-dimensional objects, the factors contributing to errors in calibration processing are analyzed systematically. The key of improving the accuracy lies in distortion correction. Combined with the characteristics of one-dimensional objects,the cross-ratio constant method is proposed to correct the image distortion. At the same time, the problems need to be paid attention in practice are pointed out, in order to enhance the algorithm stability. Experiments show that the calibration accuracy based on one-dimensional objects can be achieved at the same level of calibration based on plane, thus verifies the effectiveness of the algorithm.

关 键 词：标定 一维标定物 高精度 畸变校正 

分 类 号：O435[机械工程—光学工程]