基于自由运动一维标定物的多摄像机参数标定方法与实验  

Calibration method and experiments of multi-camera's parameters based on freely moving one-dimensional calibration object

作　　者：付强[1] 全权[1,2] 蔡开元[1] 

机构地区：[1]北京航空航天大学自动化科学与电气工程学院,北京100191 [2]北京航空航天大学虚拟现实技术与系统国家重点实验室,北京100191

出　　处：《控制理论与应用》2014年第8期1018-1024,共7页Control Theory & Applications

基　　金：国家"973"计划资助项目(2010CB327904);国家自然科学基金资助项目(61104012)

摘　　要：一维标定方法易于实现且标定效率高,为了克服现有一维标定方法的一些不足,本文提出一种用一维标定物标定多摄像机内外参数的方法,首先进行两两标定,在此过程中,假定主点坐标近似已知而仅考虑畸变、焦距、旋转和平移等参数,接着利用基本矩阵及一维标定物上特征点之间的几何约束,估计两摄像机的内外参数,两两标定完成后,采用Dijkstra最短路径法和捆绑调整对多摄像机系统进行全局标定(含主点坐标),仿真和真实实验表明本文的方法是切实有效的.The one-dimensional （1D） calibration method is of easy implementation and with high efficiency.In order to overcome deficiencies of existing 1D calibration methods,a new method is proposed in this paper to calibrate the internal and extemal parameters of multiple cameras by using a 1D calibration object.Firstly,the pair-wise calibration is carried out,where the positions of principal points are assumed to be approximately known and only those parameters such as the distortion,focal length,relative rotation and translation are needed to be taken into consideration.Then,with the fundamental matrix and geometry constraints of feature points on the calibration wand,the internal and external parameters of binocular cameras can be determined.After the pair-wise calibration,all parameters （including the positions of principal points） of the multicamera system are refined by using Dijkstra＇s shortest path algorithm and the bundle adjustment method.The simulation and real experiments demonstrated the effectiveness of the proposed method.

关 键 词：多摄像机 内外参数 标定 一维标定物 自由运动 

分 类 号：TN948.41[电子电信—信号与信息处理]