Pixel & feature level multiresolution image fusion based on fuzzy logic

Abstract

The motivation behind fusing multi-resolution images is to create a single image with improved interpretability. In algorithm (based on pixel and feature level) presented in this paper, images are first segmented into regions with fuzzy clustering and are then fed into a fusion system, based on fuzzy "if-then" rules. Fuzzy clustering offers more flexibility over traditional strict clustering; thus allowing more robustness as compared to other segmentation techniques (e.g. K-means clustering algorithm). A recently proposed subjective image fusion quality evaluation measure known as IQI (Image Quality Index) [1] is used to measure the quality of the fused image. Results and conclusion outlined in this paper would help explain how well the proposed algorithm performs. © 2007 Springer.