Fusion of multispectral and panchromatic images based on entropy and fruit fly optimization

Abstract

Image fusion (pan-sharpening) aims to generate a single image from both a panchromatic (PAN) image and multispectral images (MS). The pan-sharpened images ought to be identical to the MS images in terms of spectral information and ought also to be similar to the PAN image in terms of spatial information. Different fusion methods and algorithms have been purposed in the literature such as intensity-hue-saturation (IHS), wavelet transform (WT), principal component analysis (PCA) and Brovey transform (BT), etc. These techniques can produce color distortions in the fused images. This problem is principally due to the fact that the same details extracted from the PAN image are injected into each band of the MS images. FUFSER method utilizes the spectral response functions and Fourier transform (FT) to make an injection model. A new fusion method based on FUFSER method is presented in order to improve the spatial and spectral qualities of the fused images. This method usees local and global parameters to compute the amount of spatial details extracted from the PAN image to be added into each band of the MS images. The global parameters are computed using the fruit fly optimization, whereas the local parameters are computed using the entropy. The proposed method is applied to Plé iades and IKONOS images and compared with some existing fusion methods. The results obtained showed that the proposed method has better performance compared than other methods in terms of spatial and spectral information.