Takagi-Sugeno Fuzzy System and MTF-based Panchromatic Sharpening

摘要

The panchromatic sharpening or pansharpening refers to the fusion process of high-resolution panchromatic image and low- resolution multi-spectral images. Modulation Transfer Function (MTF) of satellite sensors has also been used for pansharpening. We investigate the use of Takagi-Sugeno fuzzy systems in MTF-based pansharpening algorithms. Traditional pansharpening schemes can result in spatial and/or spectral distortion during the fusion process. A fuzzy integrated fusion scheme is proposed to overcome this limitation. Spectral dissimilarities between panchromatic and multi-spectral bands are also taken into account. While preserving low-resolution multi-spectral information, Takagi-Sugeno fuzzy is introduced to inject appropriate spatial details in the pansharpened image. The local features of panchromatic image are also exploited to preserve the spatial and spectral content. Experiments conducted on Pliades, Spot-5 and WorldView-2 data set demonstrate the superior fusion quality of the proposed scheme.