A Novel Remote Sensing Image Fusion Algorithm Using ICA Bases

Image fusion is the method of gathering information from images obtained from sensors of different modalities in order to create an image that is more detailed than the input images. This paper presents a novel method for fusion of low-resolution multispectral (MS) images with a high-resolution panchromatic (PAN) image using Stationary Wavelet Transform (SWT) and Independent Component Analysis (ICA). Here, the MS image is modified using the PAN image at the wavelet decomposition level, followed by performing fast ICA algorithm on the individual bands, and thereby providing better resolution to the output image. In this paper, we also use parameters like correlation coefficient (CC), peak signal-to-noise ratio (PSNR), mean structural similarity measure (MSSIM) and edge stability mean square error (ESMSE) to measure the quality of the fused image. From experimental results, we observed that SWT method efficiently preserves spectral information, while ICA efficiently preserves spatial information. The proposed method is found to show better performance compared to other standard fusion methods like IHS fusion, PCA fusion, DWT fusion, SWT fusion and ICA fusion.