We have developed a new computer program for tissue segmentation designed for the quantification of smooth muscle in the corpus cavernosum. The program uses digitalized images of stained histological sections taken with a CCD camera. The recognition of the different tissues is based on the examination of absorption of monochromatic light. The section is successively illuminated with light of three different wavelengths and the three absorption values and texture parameters are used for the identification of the tissue using a fuzzy algorithm.