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". 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