Color Constancy using Superpixel

Authors: Xingsheng Yuan, Zhengzhi Wang

Abstract:

Color constancy algorithms are generally based on the simplified assumption about the spectral distribution or the reflection attributes of the scene surface. However, in reality, these assumptions are too restrictive. The methodology is proposed to extend existing algorithm to applying color constancy locally to image patches rather than globally to the entire images. In this paper, a method based on low-level image features using superpixels is proposed. Superpixel segmentation partition an image into regions that are approximately uniform in size and shape. Instead of using entire pixel set for estimating the illuminant, only superpixels with the most valuable information are used. Based on large scale experiments on real-world scenes, it can be derived that the estimation is more accurate using superpixels than when using the entire image.

Keywords: color constancy, illuminant estimation, superpixel

Digital Object Identifier (DOI): doi.org/10.5281/zenodo.1327666

References:

[1] A Gijenij, T Gevers, J v d Weijer, "Computational color constancy: Survey and Experiments," Trans. Image Processing, 2011, vol. 20, pp.2475-2489, sept. 2011.
[2] G. D. Finlayson, S. D. Hordley, and P. M. ubel, "Color by correlation: a simple, unifying framework for color constancy," IEEE Trans. Pattern Analysis and Machine Intelligence, vol.23, pp.1209-1221, Nov.2001.
[3] G. D. Finlayson, S. D. Hordley, I. astl, "Gamut constrained illuminant estimation," Int.J. Computer Vision, vol.67, no.1, pp.93-109, 2006.
[4] E. H. Land, "The retinex theory of color vision," Scientific American, vol.237, pp.108-128, Dec. 1977.
[5] G. Buchsbaum, "A spatial processor model for object colour perception," Journal of the Franklin Institute, vol.310, pp.1-26, July 1980.
[6] A. Gijsenij, and T Gevers, "Color constancy using image regions," in Conf.Rec.2007 IEEE Int Conf. Image Processing, pp.501-504.
[7] L Alex, S Adrian, Kiriakos N. Kutulakos, David J. Fleet, Sven J. Dichinson and Kaleem Siddiqi, "TurboPixels: Fast superpixels Using Geometric Flows," IEEE Trans. Pattern Analysis and Machine Intelligence, vol..31, pp.2290-2297, Dec.2009.
[8] S. A. Shafer, "Using color to separate reflection components," Color Research and Application, vol. 10, pp. 210-218, Dec.1985.
[9] H. L Edwin and J McCann., "Lightness and retinex theory," J. Opt. Soc Am, vol.61, pp.1-11, Jan. 1971.
[10] J v d Weijer, T Gevers, and A Gijsenij, "Edge-based color constancy," Trans. Image Processing, vol.16, pp.2207-2214, Sept. 2007.
[11] F Ciurea and B.V. Funt, "A large image database for color constancy research," in Proc. Eleventh Color Imaging Conf. IS&T- The Society for Imaging Science and Technology, Scottsdale, Arizona, 2003, pp. 160-164.
[12] S.D. Hordley and G.D. Finlayson, "Reevaluation of color constancy algorithm performance," J. Opt. Soc Am, vol. 23, pp. 1008-1020, Nov. 2006.