In the phenology of pixels, as the end-member proportions of a pixel vary with the progression of seasons, the pixel changes through different versions. In the image fusion of a low spatial resolution multiresolution (MS) pixel, multiple high spatial resolution fused pixels are generated. The original MS pixel and each fused pixel superimpose over different panchromatic (PAN) pixels and have different end-member proportions. Since analogies exist between pixel phenology and image fusion, the spectral change directions of pixels in phenology can be used as a reference to obtain optimal spectral change directions for MS sub-pixels in image fusion. Regarding pixel phenology, it was found that the optimal spectral change direction for an MS sub-pixel in image fusion is along the sub-pixel vector minus a haze vector. Based on this direction and a multivariate regression between the MS and PAN images, we propose a new method for image fusion. In an evaluation using spatially degraded IKONOS MS and PAN images, this method outperforms some selected current image fusion methods.
No Reference information available - sign in for access.
No Citation information available - sign in for access.
No Supplementary Data.
No Article Media
Document Type: Research Article
Department of Earth and Space Science and Engineering, York University, Toronto, Ontario, Canada
Department of Earth and Space Science and Engineering, York University, Toronto, Ontario, Canada,State Key Lab of Geological Processes and Mineral Resources, China University of Geosciences, Wuhan, China
Publication date: March 1, 2011
More about this publication?