Novel Techniques for Color Texture Classification.
Rubén Múñiz, José Antonio Corrales · 2006
Abstract: In the computer vision domain, color has not been a relevant field of study, since grayscale images contained enough information to solve many different tasks. Another reason to avoid color images, was the fact that they require to upgrade the input hardware (mainly CCD cameras) and that the CPU processing power need to be higher to be able to handle the additional color information. In the recent years, many researchers have begun to take color information into consideration. In the texture analysis field, many classical feature extraction algorithms have been enhanced to process color textures and new ones have been researched. In this paper, a new approach to extend grayscale texture analysis methods is presented. By means of the band ratioing technique, we can modify any feature extraction algorithm to take advantage of color information and achieve higher classification rates. To prove this extreme, three standard techniques has been selected: Gabor filters, Wavelets and Cooccurrence Matrices. For testing purposes, 30 color textures have been selected from the Vistex database. We will perform a number of experiments on that texture set, combining different ways of adapting the former algorithms to process color textures and extract features from them.