Multisensor image fusion by frequency spectrum substitution: subband and multirate approaches for a 3:5 scale ratio case

This work proposes two general and formal solutions to the problem of multisensor image data fusion, and presents results on a specific case study concerning multispectral observations. The former method exploits the generalized Laplacian pyramid (GLP), which is a non-dyadic bandpass analysis structure unconstrained from the ground scales of the imaged data, to transplant high spatial frequency components from one image to another. The latter methods relies on uniform subband decomposition (USD) to allow spatial frequencies spectrum substitution as well. Results on fusion of Landsat TM (30 m resolution) and MOMS-2P (18 m) are presented and discussed.