Use of band ratioing for building illumination independent texture classification systems

When a researcher faces the problem of developing a computer vision application there are many problems that have to be solved. Among those problems, we find the study of the proper illumination conditions of the scene. The way the objects are illuminated is so critical that it affects all the elements in the analysis chain. Generally the lights are carefully chosen and adapted for the computer vision system, but in some cases the system will run under variable conditions. In those cases it will be necessary to improve the robustness of the algorithms in order to minimize the impact on the application. Texture classification systems are also vulnerable to this particularity. On such systems the effect of light changes on the measured features may eventually lead to texture misclassification and performance degradation. In this paper we present an approach that will be helpful to overcome such problems when the light comes from a directional source, such as halogen projectors, LED arrays, etc.