基于摄影测量的非重叠视场相机全局标定  

Global Calibration of Non-Overlapping Field of Views Cameras Based on Photogrammetry

作　　者：胡茂邦 夏仁波[1] 陈松林[1,2] 张洪瑶 HU Mao-bang;XIA Ren-bo;CHEN Song-lin;ZHANG Hong-yao(Shenyang Institute of Automation,Chinese Academy of Sciences,Shenyang 110016,China;University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]中国科学院沈阳自动化研究所,沈阳110016 [2]中国科学院大学,北京100049

出　　处：《组合机床与自动化加工技术》2018年第10期89-92,共4页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：国家科技支撑计划资助项目(2015BAF08B01-01);国家自然科学基金项目(51375476)

摘　　要：在使用多组相机对大尺寸工件进行视觉测量时,相机之间不存在公共视场,相机间的高精度全局标定是视觉测量系统的关键。提出一种基于摄影测量的非重叠视场相机的全局标定方法,对不存在重叠视场的相机同时进行内参标定和全局标定。首先固定两个平面靶标,利用摄影测量求解两个固定靶标的位置关系,并将此关系作为约束条件,对两个非重叠视场相机分别采集一幅图像,求解出两个相机坐标系之间的转换矩阵,通过非线性优化求解转换矩阵的最优解,最后把所有相机的坐标系统一到同一坐标系下。试验证明,此方法具有较高的标定精度,测量误差RMS为0. 04mm,并且标定过程简单,适合现场标定。There is no common field of view between cameras when using multiple sets of cameras for the visual measurement of large-size workpieces.High precision global calibration between cameras is the key to a vision measurement system.A global calibration method based on photogrammetry is proposed,which performs internal and global calibration for cameras with no overlapping FOVs.Firstly,two planar targets are fixed,and the position of two fixed targets is solved by photogrammetry.The position between the two fixed targets is taken as the constraint condition,and an image is acquired for two non-overlapping FOVs cameras to solve the transformation matrix of cameras’coordinate systems.Through the nonlinear optimization,the optimal solution of the transformation matrix can be solved.Finally all the camera coordinate systems are converted to the same coordinate system.Experiments show that this method has high calibration accuracy,measurement error RMS 0.04mm,and calibration process is simple,suitable for on-site calibration.

关 键 词：全局标定 摄影测量 非重叠视场 固定靶标 

分 类 号：TH166[机械工程—机械制造及自动化] TG506[金属学及工艺—金属切削加工及机床]