Pan-Sharpening Based Satellite Multispectral Image Enhancement under NSC-Transform

Abstract: The project presents that pan sharpening used for satellite panchromatic and multi spectral images based Non-subsamples contourlet transformation. It contains two different approaches that are, NSCT with different levels of decomposition and NSCT with up-sampling based fusion of pixel level. NSCT is very well-organized in representing the information of directional and capturing intrinsic geometrical structures of the objects. A combination of spatial resolution (high) extracted from PAN images into the spectral resolution (high) of MS images gives both high spatial and spectral resolution pan sharpened image. This displays both spatial and spectral qualities while falling computation time. By applying up sampling after NSCT, structures and more information of the MS images are extra likely to be preserved. Hence, pan-sharpening is done by merging it with detail information offered every Pan image at the same fine level. The system simulated result represents that used method provides better resolution in this image and it also measured the performance parameters for example correlation, PSNR, SSIM and standard deviation.

Keywords: Image Enhancement, Non-sub sampled Contourlet Transform, Pan-sharpening Technique.