Filter-Based Image Fusion for Panchromatic and Multispectral Satellite Image to Improve Spectral Quality

Image fusion intends is to improve color information in a final image as well as adding edge (spatial) information to it. Along with the existing fusion algorithms, filter based fusion method are the most frequently discuss cases in fresh publications due to their skill to get better edge and color information of multispectral (MS) and panchromatic (PAN) images. Filter based fusion approach extract edge information from the PAN image and injects it into MS images. Visual and numerical assessment show that the proposed algorithm plainly improves the fusion quality in terms of correlation coefficient (CC), relative dimensionless global error (ERGAS) in synthesis, spectral angle mapper(SAM), universal image quality index(UIQI), and quality without reference(QNR), as compared with fusion methods, including better intensity-hue-dispersion(IHD), Bayesian, and sparse fusion of image.