A novel image fusion algorithm based on 2D EMD and IHS

Abstract

The fusion of multispectral (MS) and panchromatic (PAN) images is a useful technique to enhance the spatial quality of low-resolution MS images. The standard data fusion methods may not be satisfactory because they can alter spectral information content. In this paper, we present a new method by combining 2D EMD (two dimensional empirical mode decomposition) with IHS (Intensity-Hue-Saturation). Empirical mode decomposition (EMD) catches peoplepsilas attention for analyzing the nonstationary signals adaptively. To get better deposition results we present a modified two dimensional EMD algorithm using the Fast Radical Basis Function (FastRBF). Then we illustrate a novel image fusion algorithm based on IHS and two dimensional EMD fusion techniques. This new method combines the high frequency IMFs (intrinsic mode functions) of panchromatic image with the residual component of the intensity image. Experiments show that the new method has improved quality compared with those obtained by standard IHS, Brovey and IHS discrete wavelet transform.