A New Extension of the Rank Transform for Stereo Matching

Abstract:

Stereo matching methods often use rank transform to deal with image distortions and brightness differences prior to matching but a pixel in the rank transformed image may look more similar to its neighbor, which would cause matching ambiguity. We tackle this problem with two proposals. Firstly, instead of using two values 0 and 1,we increase the discriminative power of the rank transform by using a linear, smooth transition zone between 0 and 1 for intensities that are close together. Secondly, we propose a new Bayesian stereo matching model by not only considering the similarity between left and right image pixels but also considering the ambiguity level of them in their own image independently. We test our algorithm on both intensity and color images with brightnesss differences. Corresponding 2D disparity maps and 3D reconstruction results verify the effectiveness of our method.

• Periodical: Advanced Engineering Forum (Volumes 2-3)
• Pages: 182-187
• DOI: https://doi.org/10.4028/www.scientific.net/AEF.2-3.182
• Online since: December 2011
• Authors: Ge Zhao, Ying Kui Du, Yan Dong Tang
• Keywords: Bayesian Stereo Model, Rank Transform, Stereo Matching

References

[1] G. Gupta, M. Rawat, Region growing stereo matching method for 3D building reconstruction, International Journal of Computational Vision and Robotics , vol. 2, pp.89-98, (2011).

DOI: 10.1504/ijcvr.2011.039359

[2] K. Zhang, J. Lu, Cross-based local stereo matching using orthogonal integral images, IEEE Trans on Circuits and Systems for Video Technology, vol. 19, pp.1073-1079, (2009).

DOI: 10.1109/tcsvt.2009.2020478

[3] D. Scharstein, R. Szeliski, A taxonomy and evaluation of dense two-frame stereo correspondence algorithms, IJCV, vol. 1, pp.1917-1920, (2002).

DOI: 10.1109/smbv.2001.988771

[4] R. Zabih, J. Woodfill, Non-parametric local transforms for computing visual correspodence , ECCV, vol. 8(1), pp.151-158, (1994).

DOI: 10.1007/bfb0028345

[5] K. Ambrosh, M. Humenberger, Extending two non-parametric transforms for FPGA-based stereo matching using bayer filtered cameras, CVPR Workshop, vol. 1, pp.1-8, (2008).

DOI: 10.1109/cvprw.2008.4563146

[6] J. Banks, M. Bennamoun, Reliability analysis of the rank transform for stereo matching, IEEE Transactions on Systems, Man and Cybernetics, vol. 31, pp.870-880, (2001).

DOI: 10.1109/3477.969491

[7] K. Wang, Adaptive stereo matching algorithm based on edge detection, International Conference on Image Processing, vol. 1, pp.1-8, (2004).

DOI: 10.1109/icip.2004.1419748

[8] G. Zheng, X. Su, Local stereo matching with adaptive support-weight, rank transform and disparity calibration, Pattern Recognition Letters, vol. 1, pp.1230-1235, (2008).

DOI: 10.1016/j.patrec.2008.01.032

[9] C. Li, T. Caelli, Bayesian stereo matching , Computer Vision and Image Understanding, vol. 5, pp.85-96, (2007).

DOI: 10.1016/j.cviu.2005.09.009

[10] A. Geiger, R. Martin, R, Urtasun, Efficient large-scale stereo matching , Asian Conferencel of Computer Vision, vol. 5, pp.25-38, (2010).

DOI: 10.1007/978-3-642-19315-6_3

[11] Y. Taguchi, B. Wilburn, C. Zitnick, Stereo reconstruction with mixed pixels using adaptive over-segmentation , Computer Vision and Pattern Recognition , vol. 2, pp.1-8, (2007).

DOI: 10.1109/cvpr.2008.4587691