Moment invariants to affine transformation of colours

Most colour descriptors are not robust because they are constructed for simple colour transformations, such as a diagonal-offset transformation. In this paper, a type of colour descriptor is proposed which is composed of rational expression of moments that are invariant to full affine photometric transformations of the 3 colour bands, called a colour affine moment invariant (CAMI). The construction method, based on multiple integrations of colour cores, is intuitive, and it is generic because it can recover existing CAMIs and generate infinitely many new CAMIs as well. Through comprehensive experiments, the effectiveness of one set of newly generated CAMIs is validated. It is shown that CAMIs can better cope with changing illuminations and white balance changes than some typical colour descriptors and are moderately robust against small changes in viewpoint of 3D objects.

[1]  Filiberto Pla,et al.  Affine Illumination Compensation for Multispectral Images , 2007, SCIA.

[2]  Thomas H. Reiss,et al.  The revised Fundamental Theorem of Moment Invariants , 1991, IEEE Trans. Pattern Anal. Mach. Intell..

[3]  De Xu,et al.  Illumination-independent descriptors using color moment invariants , 2009 .

[4]  Ming-Kuei Hu,et al.  Visual pattern recognition by moment invariants , 1962, IRE Trans. Inf. Theory.

[5]  Jan Flusser,et al.  Graph method for generating affine moment invariants , 2004, ICPR 2004.

[6]  Patrick Gros.  Color illumination models for image matching and indexing , 2000, Proceedings 15th International Conference on Pattern Recognition. ICPR-2000.

[7]  Xu Wan-ying.  Moment invariants of gray distorted based on Hu moments , 2008 .

[8]  Luc Van Gool,et al.  Moment invariants for recognition under changing viewpoint and illumination , 2004, Comput. Vis. Image Underst..

[9]  Steven A. Shafer,et al.  Using color to separate reflection components , 1985 .

[10]  Luc Van Gool,et al.  Foundations of semi-differential invariants , 2005, International Journal of Computer Vision.

[11]  Jan Flusser,et al.  Pattern recognition by affine moment invariants , 1993, Pattern Recognit..

[12]  Hua Li,et al.  Gaussian-curvature-derived invariants for isometry , 2011, Science China Information Sciences.

[13]  Arnold W. M. Smeulders,et al.  The Amsterdam Library of Object Images , 2004, International Journal of Computer Vision.

[14]  Arnold W. M. Smeulders,et al.  Color Based Object Recognition , 1997, ICIAP.

[15]  Jan Flusser,et al.  Generalized affine moment invariants for object recogn , 2006, 18th International Conference on Pattern Recognition (ICPR'06).

[16]  Luc Van Gool,et al.  Vision and Lie's approach to invariance , 1995, Image Vis. Comput..

[17]  Dong Xu,et al.  Geometric moment invariants , 2008, Pattern Recognit..

[18]  Luc Van Gool,et al.  Model estimation for photometric changes of outdoor planar color surfaces caused by changes in illumination and viewpoint , 2002, Object recognition supported by user interaction for service robots.

[19]  Cordelia Schmid,et al.  Blur Robust and Color Constant Image Description , 2006, 2006 International Conference on Image Processing.

[20]  Ruixia Xu,et al.  Convex programming colour constancy with a diagonal-offset model , 2005, IEEE International Conference on Image Processing 2005.