An IHS based Multiscale Pansharpening Technique using Wavelet based Image Fusion and Saliency Detection

With the advent of remote sensing, the requirement of high resolution images that can provide high quality spectral and spatial information has increased. Due to cost and circuitry issues, the satellite sensors are capable of providing either multispectral (MS) images with high spectral and low spatial resolution or Panchromatic (pan) images with high spatial and low spectral resolution. Pansharpening of MS images is used to increase spatial resolution of MS image, using high spatial resolution pan image by fusion of MS and pan images. Several pansharpening techniques have been discussed in the literature to obtain pansharpened images with high spatial as well as spectral resolution. Among them, Intensity-Hue-Saturation (IHS), Principal Component Analysis (PCA) and Wavelet Transform based pansharpening are most commonly used techniques, as IHS and PCA provide high spatial resolution and Wavelet Transform provides high spectral resolution. In this paper, a novel pansharpening technique has been proposed which aims at integrating the advantages of IHS and Wavelet based pansharpening for providing a pansharpened image with high spatial resolution, while preserving the spectral resolution of MS image. The proposed technique is an IHS and Wavelet based multiscale image decomposition fusion technique that uses visual saliency detection for extraction of details and edge information from MS and pan images. The proposed techniques has been implemented on several image datasets containing low and very low resolution MS images and the results have been analyzed and compared with the existing techniques. The simulation results and analysis shows that the proposed pansharpening technique outperforms existing techniques by providing superior pansharpened MS images.