期刊文献+

基于计算机视觉的大型工件特征点三维坐标测量方法研究 被引量:9

The Study on 3D Coordinate Measurement of Feature Points on Large Scale Workpiece Based on Computer Vision Method
下载PDF
导出
摘要 大型工件特征点三维坐标测量具有其重要性和复杂性,近年来人们提出了各种相关的测量方法,其中计算机视觉方法以其独特的优点引起了广泛的兴趣。本文着重讨论计算机视觉方法在大型工件特征点坐标测量中的应用,提出了一种准确、快速。 The 3D coordinate measurement of feature points on large scale workpiece is important and difficult,many methods on this problem have been presented in recent years,and computer vision method has been paid more attention by researchers.This paper mainly discusses the application of computer vision method in the 3D coordinate measurement of feature points on large scale workpiece,and presents a accurate,rapid,simple,new measuring method based on the binocular stereo method and CMM. The main new idea of the method is that we replace the probe of large CMM by a camera, thus we get a noncontact 3D Coordinate Measuring Machine.Camera performs the function of aiming and measuring, binocular stereo and the complete measurement of large scale workpiece are achieved through the movement of camera. Binocular stereo method usually should solve these difficult problems such as stereo matching, the calibration of two cameras,the calibration of the distance b between two cameras and the distance v between CCD image plane and the principle point of lens.In the method,we solve these problems through the following way:Firstly,stereo matching can be avoided by extracting the feature points of left and right position images separately.Secondly,the calibration of single camera is relatively simple.Thirdly,b can be obtained through the moving length of camera,and we combine the DepthFromFocus method in monocular vision to avoid the calibration of v. An accurate,rapid,simple and practical algorithm about the accurate 2D image coordinates extracting of feature points which are always mixed with fabrication grain noise is also proposed in this paper. The experiment results indicate that the maximum absolute error in X coordinate axis is less than 0.1mm for the range of 15mm,in Z coordinate axis is less than 0.3mm for the range of 50mm, the relative error in X and Z axis are all less than 0.7%.Experiment results indicate that the correctness and reliability of this method,and it can be used in the noncontact measurement of feature points on large scale workpiece.
机构地区 哈尔滨工业大学
出处 《仪器仪表学报》 EI CAS CSCD 北大核心 1997年第6期607-612,共6页 Chinese Journal of Scientific Instrument
关键词 计算机视觉 坐标测量机 特征点 三维坐标测量 Computer Vision,CMM,Feature Points,3D Coordinate Measurement
  • 相关文献

参考文献5

二级参考文献4

共引文献50

同被引文献48

引证文献9

二级引证文献48

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部