EVALUATION OF DIFFERENT METHODS FOR USING COLOUR INFORMATION IN GLOBAL STEREO MATCHING APPROACHES

Global algorithms currently represent the state-of-the-art in dense stereo matching. These methods first set up an energy function. The energy function is then subject to optimization, which is typically achieved via graph-cuts or belief propagation. In this paper, we concentrate on the energy modelling aspect. An experimental study that focuses on the role of colour in stereo energy functions is presented. We evaluate the performance of various forms for using colour and compare it against grey-scale matching. Colour is thereby represented in nine different colour systems. The L1 andL2 distances are evaluated for computing the colour differences in the selected systems. We embed the resulting energy functions into two stereo algorithms and test them on 30 ground truth test image pairs. The results of our benchmark show that colour information, in general, leads to a significant performance gain over using intensity only. According to our evaluation results, the selection of the applied colour space is of specific importance in global stereo matching.

[1]  Masatoshi Okutomi,et al.  Color stereo matching and its application to 3-D measurement of optic nerve head , 1992, [1992] Proceedings. 11th IAPR International Conference on Pattern Recognition.

[2]  Andreas F. Koschan.  Dense Stereo Correspondence Using Polychromatic Block Matching , 1993, CAIP.

[3]  Carlo Tomasi,et al.  A Pixel Dissimilarity Measure That Is Insensitive to Image Sampling , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[4]  Olga Veksler,et al.  Fast approximate energy minimization via graph cuts , 2001, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[5]  D. Scharstein,et al.  A Taxonomy and Evaluation of Dense Two-Frame Stereo Correspondence Algorithms , 2001, Proceedings IEEE Workshop on Stereo and Multi-Baseline Vision (SMBV 2001).

[6]  Nanning Zheng,et al.  Stereo Matching Using Belief Propagation , 2002, IEEE Trans. Pattern Anal. Mach. Intell..

[7]  H. Mayer.  ANALYSIS OF MEANS TO IMPROVE COOPERATIVE DISPARITY ESTIMATION , 2003 .

[8]  Alain Crouzil,et al.  Colour Correlation-based Matching , 2005, Int. J. Robotics Autom..

[9]  Yair Weiss,et al.  Globally optimal solutions for energy minimization in stereo vision using reweighted belief propagation , 2005, Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1.

[10]  Ruigang Yang,et al.  How Far Can We Go with Local Optimization in Real-Time Stereo Matching , 2006, Third International Symposium on 3D Data Processing, Visualization, and Transmission (3DPVT'06).

[11]  Richard Szeliski,et al.  A Comparative Study of Energy Minimization Methods for Markov Random Fields , 2006, ECCV.

[12]  Heiko Hirschmüller,et al.  Evaluation of Cost Functions for Stereo Matching , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[13]  Margrit Gelautz,et al.  Simple but Effective Tree Structures for Dynamic Programming-Based Stereo Matching , 2008, VISAPP.

[14]  W. Cao,et al.  The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol. XXXVIII-4/C7 CLASSIFICATION OF HIGH RESOLUTION OPTICAL AND SAR FUSION IMAGE USING FUZZY KNOWLEDGE AND OBJECT-ORIENTED PARADIGM , 2010 .