Inter Intensity and Color Channel Co-occurrence Histogram for Color Texture Classification
In this paper we propose a new method to analyze the color texture image based on inter intensity and color channel co-occurrence histogram, which characterizes the color texture more effectively. This corresponds to the relationships between intensity and color channel along with their neighboring pixels. The proposed color texture descriptor is experimented on VisTex texture dataset. The results are analyzed and compared with Local Binary Patterns (LBP) method and Histogram ratio method. The computational intelligence-based approach, namely, fuzzy classification is used for texture classification. The proposed descriptors achieve better classification results when compared with other two methods. The proposed color texture descriptors are sufficiently robust and precise to distinguish images of different textures even if the sample size is small. The results suggest that the proposed color texture descriptors have the potential for use in real-world applications involving recognition of patterns in digital images.
KeywordsColor texture Texture classification Inter co-occurrence histogram Color-channel Inter intensity and color
We are thankful to referees for their helpful comments and suggestions.
- 19.Maenpaa, T., Pietikainen, M., Viertola, J.: Separating color and pattern information for color texture discrimination. In: Proceedings of the 16th IEEE International Conference on Pattern Recognition, vol. 1, pp. 668–671 (2002)Google Scholar
- 20.Rosenfeld, A., Wang, C.Y., Wu, A.Y.: Multispectral texture. IEEE Trans. Syst. Man Cybern. SMC-12(1), 79–84 (1982)Google Scholar
- 22.VisTex: Vision texture database of MIT media lab (1995). http://www-white.media.mit.edu/vismod/imagery/VisionTexture/vistex.html
- 23.Cateni, S., Colla, V., Vannucci, M., Borselli, A.: Fuzzy inference systems applied to image classification in the industrial field. In: Fuzzy Inference System-Theory and Applications. InTech (2012)Google Scholar