The Ehlers fusion method, which combines a standard intensity-hue-saturation (IHS) transform with the fast Fourier transform (FFT) filtering, is a good color preservation method for multi-temporal and multi-sensor datasets. However, for the dataset with more than three bands, the fusion process is complicated. Because only every three bands are fused repeatedly for multiple times until all bands are fused. The hyperspherical color sharpening (HCS) fusion method can fuse the dataset with more than three bands. The HCS approach uses a transform between n-dimensional Cartesian space and n-dimensional hyperspherical space. Moreover, from a structural point of view, the hyperspherical color space is very similar to the IHS color space. Although the existing principal component analysis (PCA) and Gram Schmidt (GS) methods are also able to integrate the dataset with more than three bands, but they, especially the PCA, usually lead to the spectral distortion. Hence, we propose the Ehlers-HCS fusion method to fuse n band datasets, even to hyperspectral images. The original Ehlers fusion renamed as Ehlers-IHS fusion for easy distinction. The WorldView-2 dataset including a panchromatic and eight multispectral bands are used as the input experimental real data. The fused images are visually and quantitatively analyzed for the spectral preservation and spatial improvement to show the effectiveness of the Ehlers-HCS fusion.
展开▼
