On the use of Image Data Information for getting a Brightness’ Perceptual Fuzzy Model

The basic difference between human and machine vision systems may be briefly summarized as "humans recognize, machines measure" [1]. However, it is not possible to simply transfer or copy the capability of human biological vision to machine vision systems. So, it is a very difficult task to design an artificial system to segment an image, based on the information supplied by the gray-level of the pixels of the image. The gray-level of the pixels within an image displays a high variability level, and it is not homogeneous across the image. Moreover, typical effects as noise, and non-uniform illumination have to be considered. Aforementioned capability of human vision system, for recognizing and discriminating brightness levels, has been considered in present work for the design of a fuzzy vision model which improves the performance of Brightness feature within image processing systems. Based on a set of rules and the data provided by the image, the proposed model derives a fuzzy set for each Brightness label.