A Study Of Multiband Image Fusion Based On Intuitionistic Fuzzy Set

Multi-band images collected by the multi-band high precision detection system can obtain more accurate and complete information through image fusion.However,due to the different characteristics of imaging equipment and image transmission,as well as the influence of different illumination,the collected images will inevitably introduce uncertain information such as noise and blur.In order to solve the fuzzy uncertainty existing in the fusion of complex detection images and the current problem of pairwise sequential fusion,this paper studies the multi-band image fusion method by using the intuitionistic fuzzy set theory which has the ability to deal with uncertainty and nonlinearity.The main work is as follows:(1)A comparative study of intuitionistic fuzzy set in multi-band image fusion.In order to solve the problem of uncertain performance of existing intuitionistic fuzzy set methods for multi-band images,five intuitionistic fuzzy set methods are used to intuitionistic fuzzify the multi-band images,and the same fusion rules are used for fusion.Through experimental simulation and evaluation index analysis,the intuitionistic fuzzy set suitable for multi-band image fusion is optimized.(2)An adaptive fusion rule based on ranking of intuitionistic fuzzy sets is proposed.In order to solve the problem of inaccurate assignment of attribute weights in image fusion,an adaptive fusion method based on intuitionistic fuzzy set ranking is proposed.Firstly,the image is divided into foreground and background by threshold segmentation method.The method of intuitionistic fuzzy set selected by work(1)is used to fuzzify the foreground.The enhanced image is obtained by combining with the background.Secondly,the enhanced image is decomposed into high and low frequency images by multi-scale transformation method.High frequency are fused using maximum absolute value.Low frequency are fused using intuitionistic fuzzy set ranking adaptive fusion rules.Finally,the fused image is obtained by inverse transform of high and low frequencies.This method effectively fuses the target information of the image,and the fusion result has rich detail information,clear edge,enhanced brightness and contrast,and improves the visual effect of the image.(3)Design and implementation of multi-band image fusion system.Design and development of the whole multi-band image fusion system based on MATLAB.The main functions are intuitionistic fuzzification,image fusion and image evaluation.Multi-band image fusion based on intuitionistic fuzzy sets is realized.