Fast panchromatic sharpening for high-resolution multi-spectral images

For IKONOS imagery, both multi-spectral and panchromatic data are provided with different spatial resolutions. In order to produce multi-spectral images with high spatial resolution, a number of panchromatic sharpening methods have been developed. However, the classical methods such as IHS, PCA and Brovey transform often introduce severe color distortion due to injecting undesired low-pass component of panchromatic observation into the fused multi-spectral images. Recent methods employing the discrete wavelet transform (DWT) or the undecimated DWT preserve good spectral information and prevent adding low-pass components. However, it is difficult to process a large volume of data due to memory requirement and computational cost. The high-pass filtering, Laplacian pyramid, and high-pass modulation (HPM) methods provide good performance with a few spectral distortion and low computational complexity. In this paper, we present a fast and general panchromatic sharpening method for combining low-resolution multi-spectral images with a high-resolution panchromatic image. Experimental results demonstrate that the proposed method produces high-quality fused image and outperforms existing panchromatic sharpening methods in terms of objective quality measures such as UIQI, Q4, RASE and ERGAS.