DETERMINE SIMILARITY OF FACES USING SYMBOLS AND AVERAGE COLOR SPECTRUM CHARACTERISTICS WITH FUZZY SYSTEM

One of the most widely used areas in the field of image processing is face recognition based on appearance characteristics, in this study to determine its similarity to the face, first we capture 200 samples of images of different people as RGB color mode and we convert this images to HSV color mode and we quantized HSV color mode to 32 levels, then we obtains the symbols of each image with the average color of each symbol. We calculate desired parameters based on angles places being created symbols and average color spectrum and stored in our database. And then based on the angle between the symbols used in the files, the desired level and parameters of color symbol Fuzzy set of files and ultimately the level of similarity are obtained based on defined fuzzy rules. Based on the observed experimental results, this method compared to determine similarity based on the symbol is 12 percent more efficient.