IMPLEMENTATION OF A PRACTICAL METHODOLOGY OF IMAGE SEGMENTATION AND BINARIZATION USING THE IMAGE-PRO PLUS 6.0 COMMERCIAL SOFTWARE

Abstract

This work described in details a set of experimental procedures implemented for a specific case, aiming at the accomplishment of the steps of gray scale images segmentation and binarization, using for this the Image-Pro Plus 6.0 (Media Cybernetics, Silver Spring, USA), a traditional and commercial image processing software. For this end, the images of the two sides of a coin of R$ 0.25 (25 cents) were acquired and then the procedures needed for segmentation and binarization of these images were performed in four steps, i.e., the conversion of the original images to gray scale object, two filters/masks application steps, and one step of segmentation by thresholding. To perform the conversion to gray scale it was chose the gray scale class 8, which generated new images with 8 bits and 256 shades of gray, a size considered light and suitable for additional processing steps. In the methodology adopted, one of the filtering step was done before the thresholding, using the flatten filter, which performed a process of the image pixels uniformization, thus contributing to reduce the intensity variations of the background pixels relative to the pixels of the objects of interest. On the other hand, the median filter, which was applied after the thresholding, acted reducing the degree of noise in the image, and consequently, making the binary object more sharpness and defined.