This paper shows the results obtained after classifying 75 cells of the vaginal epithelium, using operations of preprocessing, segmentation and representation of digital images. An application software was structured to analyze the classification parameters in the cells, such as the relationship between nucleus – cytoplasm, tint, saturation, luminance, form and position of the nucleus, in order to identify the superficial, intermediate, basal, parabasal and displasic cells (these are present in the uterus cancer and other alterations of cervix). It was possible to classify 67 cells, representing 89% of applicability skill, and 11% of error.