构建虚拟立体靶标的大视场高精度视觉标定
发布时间：2019-06-20 12:21
【摘要】：为了提高立体视觉系统在大视场下的测量精度,基于误差溯源思想提出了一种构建虚拟立体靶标的大视场高精度视觉系统标定方法,克服了大尺寸高精度标定物难以制造等问题。对影响立体视觉系统测量精度的主要因素进行分析,列出视觉测量系统的误差溯源链,解析了大视场视觉系统精度瓶颈的原因。借助激光跟踪仪,运用非线性最小二乘单位四元数算法求解坐标系刚体变换,获取大范围高精度的空间点阵,构建虚拟靶标。在相机畸变模型中考虑了三阶径向畸变和二阶切向畸变参数,并使用Levenberg-Marquardt迭代算法进行标定参数求解,进一步提高系统精度。实验构建了一套测量空间约为4m×3m×2m的双目立体视觉系统,通过对某型号高精度直线导轨进行点距测量,在测量距离3m处,152组不同长度的横向距离测量的误差算术均值为-0.003mm,误差标准差为0.08mm。测量精度相较于传统的平面标定法有较大提升。
[Abstract]:In order to improve the measurement accuracy of stereo vision system under large field of view, a calibration method of large field of view and high precision vision system for constructing virtual stereo target is proposed based on the idea of error traceability, which overcomes the problem that it is difficult to manufacture large size and high precision calibration object. The main factors affecting the measurement accuracy of stereo vision system are analyzed, the error traceability chain of visual measurement system is listed, and the reasons for the bottleneck of precision of vision system with large field of view are analyzed. With the help of laser tracker, the nonlinear least square unit quaternion algorithm is used to solve the rigid body transformation of coordinate system, and a large range and high precision spatial lattice is obtained, and the virtual target is constructed. The third-order radial distortion and second-order tangential distortion parameters are considered in the camera distortion model, and the calibration parameters are solved by Levenberg-Marquardt iterative algorithm to further improve the accuracy of the system. A binocular stereo vision system with a measuring space of about 4m 脳 3m 脳 2m is constructed experimentally. by measuring the point distance of a certain type of high precision linear guideway, at the measuring distance of 3m, the arithmetic mean error of 152 groups of transverse distance measurement of different lengths is-0.003mm, and the error standard deviation is 0.08mm. Compared with the traditional plane calibration method, the measurement accuracy is greatly improved.
【作者单位】： 上海大学机电工程及自动化学院;
【基金】：国家自然科学基金面上项目(No.41376169,No.51205243) 国家重点研发计划重点专项课题(No.2016YFC0302402)
【分类号】：TP391.41
[Abstract]:In order to improve the measurement accuracy of stereo vision system under large field of view, a calibration method of large field of view and high precision vision system for constructing virtual stereo target is proposed based on the idea of error traceability, which overcomes the problem that it is difficult to manufacture large size and high precision calibration object. The main factors affecting the measurement accuracy of stereo vision system are analyzed, the error traceability chain of visual measurement system is listed, and the reasons for the bottleneck of precision of vision system with large field of view are analyzed. With the help of laser tracker, the nonlinear least square unit quaternion algorithm is used to solve the rigid body transformation of coordinate system, and a large range and high precision spatial lattice is obtained, and the virtual target is constructed. The third-order radial distortion and second-order tangential distortion parameters are considered in the camera distortion model, and the calibration parameters are solved by Levenberg-Marquardt iterative algorithm to further improve the accuracy of the system. A binocular stereo vision system with a measuring space of about 4m 脳 3m 脳 2m is constructed experimentally. by measuring the point distance of a certain type of high precision linear guideway, at the measuring distance of 3m, the arithmetic mean error of 152 groups of transverse distance measurement of different lengths is-0.003mm, and the error standard deviation is 0.08mm. Compared with the traditional plane calibration method, the measurement accuracy is greatly improved.
【作者单位】： 上海大学机电工程及自动化学院;
【基金】：国家自然科学基金面上项目(No.41376169,No.51205243) 国家重点研发计划重点专项课题(No.2016YFC0302402)
【分类号】：TP391.41
【相似文献】
相关期刊论文 前10条
1 徐静;;基于非线性模型的摄像机标定技术研究[J];现代电子技术;2008年12期
2 李小峰;李峰;;单目摄像机标定方法的研究[J];计算机工程与应用;2009年15期
3 牟雅丹;;双目立体视觉系统的摄像机标定技术[J];科技资讯;2009年33期
4 林海平;陈静;华艳秋;;一种新型的摄像机标定方法[J];电气自动化;2009年05期
5 袁杰;朱劲;王晓年;;分布式环境下的多摄像机标定[J];电脑知识与技术;2010年02期
6 袁理;;基于平面靶标的线性摄像机标定[J];科技创新导报;2010年07期
7 高忠国;张洪;王露露;曹毅;;一种基于智能优化的摄像机标定方法研究[J];包装工程;2011年21期
8 王建华;冯帆;梁伟;王惠萍;;一种摄像机标定系统的设计与实现[J];电脑知识与技术;2012年03期
9 陈敏;;摄像机标定方法的建模与仿真研究[J];计算机仿真;2012年03期
10 臧希U，
本文编号：2503216
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2503216.html
