|
Article information
2017 , Volume 22, ¹ 4, p.80-94
Panin S.V., Chemezov V.O., Lyubutin P.S.
Determination of optical flow by complex algorithm on weighted full search
One of the important applied directions of machine vision is related to solving problems of experimental mechanics and particularly estimating three-dimensional deformation through analysis of image sequences or videostreams. The operating principle for the systems of this kind is underlined by the technique of digital image correlation (DIC). However, the algorithms employed in the commercially available systems are not sufficiently accurate for determination of displacements at the edges of the objects. This results in difficulties in construction for a three-dimensional model of the working scene in 3D vision. In doing so, the problem related to the development of optical flow search algorithms in the three-dimensional computer vision systems is actual both in scientific and applied areas. To improve the accuracy of determination for an optical flow, both at the edges of objects, as well as over the entire image it was suggested to combine the best approaches based on local analysis of images as well as “patch-match” technique. Local methods were chosen since they ensure the opportunity to perform parallel calculations. In doing so, the motion estimation algorithm consists of a hierarchical search of weighted “patch-match” function with attracting more information channels obtained by preprocessing of input images followed by sub-pixel precision approximation at each hierarchical level and the final filtration of the obtained optical flow. Test results of the proposed algorithm have demonstrated its competitiveness. The latter possessed the accuracy of the optical flow determination being comparative to the best competitors with a clear advantage to determine the optical flow at the edges of objects.
[full text]
Keywords: image processing, optical flow, motion estimation, patch match, edges of objects, end point error, weighted block difference function, complex algorithm, subpixel accuracy
Author(s):
Panin Sergey Viktorovich
Dr. , Professor
Position: Head of Laboratory
Office: Institute of Strength Physics and Materials Science of SB RAS, National Research Tomsk Polytechnic University
Address: 634021, Russia, Tomsk
Phone Office: (3822)286-904
E-mail: svp@ispms.tsc.ru
Chemezov Vitaly Olegovich
Position: Student
Office: Institute of Strength Physics and Materials Science of SB RAS
Address: 634021, Russia, Tomsk
Phone Office: (3822)286-889
E-mail: vpointc@rambler.ru
Lyubutin Pavel Stepanovich
PhD.
Position: Junior Research Scientist
Office: Institute of Strength Physics and Materials Science of SB RAS
Address: 634021, Russia, Tomsk
Phone Office: (3822)286-889
E-mail: psl@sibmail.com
References:
[1] Schreier, H., Orteu, J., Sutton, M.A. Image correlation for shape, motion and deformation measurements. Basic concepts, theory and applications. New York: Springer Science; 2009: 321.
[2] LaVision Techniques of digital image correlation. Available at: http://www.lavision.de/en/techniques/dic dvc/index.php. (accessed 25.06.2016).
[3] Fortun, D., Bouthemy, P., Kervrann, C. Optical flow modeling and computation: A survey. Computer Vision and Image Understanding. 2015; (135):1–21.
[4] Horn, B., Schunck, B. Determining optical flow. Artificial Intelligence. 1981; (16):185–203.
[5] Sundaram, N., Brox, T., Keutzer, K. Dense point trajectories by GPU-accelerated large displacement optical flow. Lecture Notes in Computer Science. 2010; (6311):438-451.
[6] Baker, S., Scharstein, D., Lewis, J., Roth, S., Black, M.J., Lewis, J., Roth, S., Szeliski, R. A database and evaluation methodology for optical flow. Proceedings of the IEEE International Conference on Computer Vision (ICCV), Rio de Janeiro, Brazil, October 2007. Institute of Electrical and Electronics Engineers ( IEEE ); 2007:1-8.
[7] Baker, S., Scharstein, D., Lewis, J., Roth, S., Black, M., Szeliski, R. A database and evaluation methodology for optical flow. Computer Vision. 2011; 92(1):1–31.
[8] Barnes, C., Shechtman, E., Finkelstein, A., Goldman, D. PatchMatch: a randomized correspondence algorithm for structural image editing. ACM Transactions on Graphics (Proceedings SIGGRAPH). 2009; 28(3). Article No. 24.
[9] Z. Chen, H. Jin, Z. Lin, S. Cohen, and Y. Wu Large displacement optical flow with nearest neighbor field. Proceedings of the 2013 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW '13), June 23 - 28, 2013. Washington, DC, USA: IEEE Computer Society; 2013: 2443–2450.
[10] Kim, T., Lee, H., Lee, K. Optical flow via locally adaptive fusion of complementary data costs. IEEE International Conference on Computer Vision (ICCV), 01 - 08 Dec, 2013. Sydney, Australia: Institute of Electrical and Electronics Engineers ( IEEE ); 2013: 3344–3351.
[11] Zimmer, H., Bruhn, A., Weickert, J., Valgaerts, L., Salgado, A., Rosenhahn, B., Seidel, H.-P. Complementary optic flow. Lecture Notes in Computer Science. 2009; (5681):207-220.
[12] Tao, M., Bai, J., Kohli, P., Paris, S. A non-iterative, sublinear optical flow algorithm // Computer Graphics Forum (Eurographics 2012). May 2012. Vol. 31, No. 2. P. 345–353. Michael W. Tao, Jiamin Bai, Pushmeet Kohli, and Sylvain Paris. "SimpleFlow: A Non-iterative, Sublinear Optical Flow Algorithm". Computer Graphics Forum (Eurographics 2012), 31(2), May 2012. P. 345–353. ýòî æóðíàë!) Computer Graphics Forum, 2011, The Eurographics Association and Blackwell Publishing Ltd. Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
[13] Itseez OpenCV. Available at: http://www.opencv.org. (accessed 25.06.2016).
[14] Werlberger, M., Pock, T., Bischof, H. Motion estimation with non-local total variation regularization. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 13-18 June, 2010. San Francisco, CA, USA: Institute of Electrical and Electronics Engineers ( IEEE ); 2010: 2464–2471.
[15] Rosenberg, Y., Werman, M. Representing local motion as a probability distribution matrix applied to object tracking. Proceedings of the 1997 Conference on Computer Vision and Pattern Recognition (CVPR '97), 17-19 June, 1997. San Juan, PR, USA: IEEE Computer Society; 1997: 654–659.
[16] Gonzalez, R.C., Woods, R.E. Digital image processing. Upper Saddle River: Prentice Hall; 2002: 793.
[17] Anandan, A. Computational framework and an algorithm for the measurement of visual motion. International Journal of Computer Vision. 1989; 2(3):283–310.
[18] Yoon, K.-J., Kweon, I. S. Adaptive support-weight approach for correspondence search. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2006; 28(4):650–656.
[19] Debella-Gilo, M., Kaab, A. Sub-pixel precision image matching for measuring surface displacements on mass movements using normalized cross-correlation. Remote Sensing of Environment. 2011; (115):130–142.
[20] Xu, L., Jia, J., Matsushita, Y. Motion detail preserving optical flow estimation. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2012; 34(9):1744–1757.
[21] Sun, D., Roth, S., Black, M.J. Secrets of optical flow estimation and their principles. The Twenty-Third IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2010, San Francisco, CA, USA, 13-18 June, 2010. USA: Institute of Electrical and Electronics Engineers (IEEE ); 2010: 2432-2439. DOI: 10.1109/CVPR.2010.5539939.
Bibliography link:
Panin S.V., Chemezov V.O., Lyubutin P.S. Determination of optical flow by complex algorithm on weighted full search // Computational technologies. 2017. V. 22. ¹ 4. P. 80-94
|
