A Multifocus Image Fusion Scheme Based on Similarity Measure of Transformed Isosceles Triangles Between Intuitionistic Fuzzy Sets
Qian Jiang, Shin‐Jye Lee, Xiao‐Jun Zeng, Xin Jin, Jingyu Hou, Wei Zhou, Shaowen Yao
Abstract
Intuitionistic fuzzy set (IFS) theory is widely used to solve imprecise and uncertain problems, and the similarity measure between IFSs can describe the similarity or dissimilarity degree of objects in the real word. In image fusion, it is a key issue to measure and extract the features of multifocus images to fuse them. This research proposes a similarity measure technique of IFSs, which is introduced into image fusion to measure the feature similarity of adjacent pixels. In this work, we present an available similarity measure method between IFSs and propose an image fusion method based on this measure and also design an edge reserved method based on the Canny edge detection to reinforce the image fusion effect. First, the proposed similarity measure is introduced and proved. Second, a focus detection method that can detect the focused regions of an image is designed based on the proposed measure method of IFSs to obtain a preliminary decision map. Third, morphological operations are employed to generate the secondary decision map. Fourth, a Canny edge model-based technique is designed to preserve the edges of the source images. Extensive experiments show that the presented similarity measure method of IFSs is reliable, and the fuzzy measure-based image fusion scheme is effective compared with current methods.