摘要
针对光笔式双目视觉测量系统的标定问题,讨论了关于相机内参、双相机外参以及测量笔的相关标定理论,开发了一整套基于LabVIEW的标定系统。运用张氏平面标定法实现了相机内参标定。结合基于标准长度的外部参数标定方法,实现了双相机外参数标定。运用粒子群算法和LM算法相结合,加快了目标函数高维寻优速度。在测量笔标定环节,提出了一种基于最小二乘法的现场校准方法。标定系统完成了后期开展相关测量前的所有准备工作,具有较高的精度和实用性。在测量系统标定结果基础上对直径25mm标准陶瓷球进行测量,测量结果标准差达到0.019mm。
Aiming at the calibration problem of the binocular vision measurement system using a light pen,the theory about the calibration of camera intrinsic parameters,double-camera extrinsic parameters and measuring pen was discussed.At the same time,the whole calibration system based on LabVIEW was developed.The camera intrinsic parameters calibration was achieved by using Zhang plane calibration method,while the double-camera extrinsic parameters calibration was achieved by the method based on standard length.Using the method combining the particle swarm optimization with the Levenberg-Marquard(LM)algorithm,the high-dimensional optimization target function was speeded up effectively.In the process of measuring pen calibration,a very convenient field calibration method based on least square was proposed.The calibration system could complete all the preparation work before the relevant measurement.It has high precision and practicability.Based on the calibration results of measurement system,the standard deviation of test result for the 25 mm diameter standard ceramic ball is 0.019 mm.
出处
《应用光学》
CAS
CSCD
北大核心
2015年第5期784-790,共7页
Journal of Applied Optics
基金
江苏省人事厅江苏省博士后资助计划(1201012C)
关键词
双目视觉
标定
外部参数
内部参数
粒子群算法
LM算法
binocular vision
calibration
extrinsic parameters
intrinsic parameters
particle swarm optimization
LM algorithm
