Abstract
<jats:p>When processing images, one of the most difficult tasks is their segmentation, which is due to the lack of a priori information about the presence, objects of interest shape and location, the presence of low-contrast areas in which the sought-after objects of interest may be partially or completely located, noise, image blur and other factors. Currently, one of the common approaches to solving this problem is various algorithms based on the applying of type-1 and type-2 fuzzy sets usage. This paper proposes to improve the sensitivity and reliability of the iterative algorithm for segmentation of grayscale images by changing the membership functions calculating method, as well as reducing the number of control parameters. Experimental results are presented using the example of segmentation of real grayscale medi-cal images.</jats:p>