首页 | 本学科首页   官方微博 | 高级检索  
     

稳定高精度的双目立体视觉测量系统标定方法
引用本文:何万涛,梁永波,李景贺. 稳定高精度的双目立体视觉测量系统标定方法[J]. 黑龙江科技学院学报, 2013, 0(5): 463-468
作者姓名:何万涛  梁永波  李景贺
作者单位:黑龙江科技大学现代制造工程中心,哈尔滨150022
基金项目:国家科技重大专项项目(2010ZX04016-012);黑龙江省教育厅科学技术研究项目(12531601)
摘    要:
为了提高和稳定双目立体视觉结构光测量系统的精度,提出一种简单有效的标定流程及基于极线几何约束优化双目立体视觉结构光测量系统外部参数的方法.采用平面板标定理论分别标定左右摄像机的内部参数,将标定的内参数作为双目标定的内参数在多视角下标定双目测量系统的外部参数.利用极线几何约束,通过最小化多视角下同一对应点的极线误差,优化计算双目测量系统的外部参数.实验结果表明,该方法计算得到的双目测量系统参数使得测量系统在整个测量空间体积内的测量精度高且稳定,在双目立体视觉测量系统的工业化检测应用中具有重要的意义.

关 键 词:双目立体视觉  结构光  测量  系统标定  极线几何

Accurate and stable calibration technique for binocular stereo measurement system
HE Wantao,LIANG Yongbo,LI Jinghe. Accurate and stable calibration technique for binocular stereo measurement system[J]. Journal of Heilongjiang Institute of Science and Technology, 2013, 0(5): 463-468
Authors:HE Wantao  LIANG Yongbo  LI Jinghe
Affiliation:1.Modern Manufacture Engineering Center, Heilongjiang University of Science & Technology, Harbin 150022, China;)
Abstract:
This paper is targeted at improving accuracy and stabilization of binocular stereo measurement system and presents a simple and effective calibration procedure and an external parameter optimizing method based on epipolar geometry.The method consists of calibrating the internal parameter of left and right camera respectively using the planar calibration method,calibrating the external parameter of the binocular stereo measurement system in every view using the parameters as initial value,and computing optimization result by minimizing epipolar line error of the same pair of points in every view.Experimental results that the method could produce binocular measurement system parameters which give the measurement system a higher and more stable precision in the whole measurement space volume and promises to play an important part in industrialized detection and application of the binocular stereo vision measurement system.
Keywords:binocular stereo  structure light  measurement system calibration  epipolar geometry
本文献已被 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号