Διαισθητική ασαφής επεξεργασία εικόνας

Abstract

The purpose of this doctoral dissertation is to establish a flexible mathematical framework for image processing, based on the concepts of intuitionistic fuzzy sets theory. In view of performing this task, the problem of analyzing and synthesizing a digital image to and from its intuitionistic fuzzy components is addressed and analytic, as well as heuristic methods are presented. The modeling is carried out by exploiting and interpreting the inherent ambiguity and vagueness, carried by the image itself, in terms of elements of intuitionistic fuzzy sets. In order to provide the necessary tools for the proposed image processing architecture, the theory of intuitionistic fuzzy sets is expanded by introducing novel theoretical concepts and definitions. The efficiency of the intuitionistic fuzzy framework is demonstrated in the context of contrast enhancement, image segmentation and edge detection, by considering images from diverse imaging modalities, including medical images. Experimental evaluation of the proposed architecture yields satisfactory and promising results compared to their fuzzy counterparts, as well as to traditional methods. Additionally, extensions to color images are also provided. Finally, the notion of complex intuitionistic fuzzy sets is introduced that allows for future application of the developed techniques to images in the frequency domain.