Vision measurement method based on Bayesian model and digital image correlation
-
摘要: 为了能够快速准确地实现3维测量，提出了一种基于贝叶斯模型和数字图像相关的视觉测量方法。该方法采用数字图像相关法对校正后的图像进行立体匹配，克服了传统立体匹配方法精度不高的问题；采用贝叶斯模型估计图像视差，并将其作为数字图像相关法非线性迭代优化的视差初值，克服粗搜索方法寻找视差初值计算量大、精度低的缺点；基于校正后图像与原图像之间的投影关系，由最小二乘法计算出匹配点的3维坐标。结果表明，基于贝叶斯模型和数字图像相关的视觉测量算法，能够快速准确地实现3维视觉测量。Abstract: In order to operate 3-D measurement efficiently and accurately, a vision measurement method correlation was proposed based on Bayesian model and digital image. Digital image correlation was used to match the rectified images, which had the ability to improve the accuracy of traditional stereo matching method. The iterative initial disparity of digital image correlation was determined by Bayesian model to overcome the disadvantages of coarse-fine search. Based on the projecting relationship between the rectified image and the raw image, 3-D coordinates of the matched point were then calculated by the least square method. The results show that the proposed vision measurement method based on Bayesian model and digital image correlation has high accuracy and efficiency of 3-D vision measurement.
-
Keywords:
- image processing /
- vision measurement /
- digital image correlation /
- Bayesian model /
- stereo matching
-
-
[1] HAN Y H. Research on 3-D model reconstruction method based on binocular stereo vision[D]. Taiyuan:North University of China, 2014:8-20(in Chinese).
[2] ROY S, COX I J. A maximum-flow formulation of the n-camera stereo correspondence problem[C]//Sixth International Conference on Computer Vision, 1998.New York,USA:IEEE, 1998:492.
[3] SUN, H Y S J, ZHENG N N. Stereo matching using belief propagation[J]. IEEE Transactions on Pattern Analysis Machine Intelligence, 2003, 25(7):787-800.
[4] SUTTON M A, MCNEIL S R, HELM J D, et al. Advances in two-dimensional and three-dimensional computer vision[J]. Photomechanics, 2000, 77(1):323-372.
[5] YU Q F, LU H W, LIU X L. Precisely measurement and motion based on image[M].Beijing:Beijing Science Press, 2002:153-154(in Chinese).
[6] GAO Y. Research on key technologies and applications of three-dimensional digital image correlation[D]. Hefei:University of Science and Technology of China, 2014:27-37(in Chinese).
[7] GEIGER A, ROSER M, URTASUN R. Efficient large-scale stereo matching[C]//TenthAsian Conference on Computer Vision-volume Part Ⅰ, 2010.Berlin,Germany:Springer-Verlag, 2010:25-38.
[8] LI M J, SU X Y. Stereo matching by means of digital speckle[J]. Laser Technology, 2004, 28(5):550-553(in Chinese).
[9] ZHANG Z. A flexible new technique for camera calibration[J]. IEEE Transactions on Pattern Analysis Machine Intelligence, 2000, 22(11):1330-1334.
[10] HEIKKILA J, SILVEN O. A four-step camera calibration procedure with implicit image correction[C]//IEEE Computer Society Conferenceon Computer Vision and Pattern Recognition, 1997.New York,USA:IEEE, 1997:1106-1112.
[11] BRADSKI G R, KAEHLER A. Learning opencv-computer vision with the opencvlibrary[M].Sebastopol,CA,USA:O'Reilly Media,Inc,2008:467-476.
[12] DEVERNAY F, FAUGERAS O. Computing differential properties of 3-D shapes from stereoscopic images without 3-D models[C]//IEEE Computer Society Conference on Computer Vision and Pattern Recognition,1994.New York,USA:IEEE,1994:208-213.
[13] VENDROUX G, KNAUSS W G. Submicron deformation field measurements:part 2. Improved digital image correlation[J]. Experimental Mechanics, 1998, 38(2):86-92.
[14] GARCIA D, ORTEU J J, PRNAZZI L. A combined temporal tracking and stereo-correlation technique for accurate measurement of 3-D displacements:application to sheet metal forming[J]. Journal of Materials Processing Technology, 2002, 125(2):736-742.
计量
- 文章访问数:
- HTML全文浏览量:
- PDF下载量:
下载: