Fuzzy Image Segmentation via Thresholding

摘要

This paper presents ah image segmentation method referred to as fuzzy image segmentation via thresholding (FIST). The threshold value that is needed for the segmentation procedure is obtained through a programmed automated computation. The histogram of the image and the measure of fuzziness constitute the first stage of the segmentation procedure whereby the threshold value is obtained. Inputting the threshold value into a 'split and merge' procedure leads to the final stage. This procedure of FIST has been applied successfully to image processing requirements for autonomous mobility of blind individuals.