Abstract
In this paper, a method for edge detection in digital images based on the interval-valued intuitionistic fuzzy concepts is presented. We propose a novel method to generate the Interval-Valued Intuitionistic Fuzzy Sets (IVIFS) from Interval-Valued Fuzzy Sets (IVFS). Given an input image in the gray level domain, first we construct a fuzzy image, and then each element of the image is converted to an interval-valued fuzzy element based on t-norms, t-conorms and the neighboring elements of that element. We then introduce an Interval-Valued Intuitionistic Fuzzy Generator (IVIFG) to construct the IVIFS for the given image. The optimized value for the negation parameter is determined based on a notion of the entropy of IVIFSs. Using this IVIF image, an intuitionistic fuzzy edge is introduced and finally this edge image is converted to a fuzzy and gray level edge image in the following stages of the proposed framework. Our experiments on a standard image database demonstrate the superiority of the proposed approach. In particular, our method shows a better performance compared to IVF case. Different t-norms, t-conorms as well as different neighboring sizes of each element of the image are used and compared.