A Robust Pan‑Sharpening Scheme for Improving Resolution of Satellite Images in the Domain of the Nonsubsampled Shearlet Transform

摘要

Pan-sharpening is a substantial post-processing task for captured multispectralremotely sensed satellite imagery. Its main purpose is to fuse the high spectral characteristicsof the multispectral (MS) images with the high spatial information ofthe panchromatic (Pan) image to output a sharper MS image (pan-sharpened) thatencompasses higher spectral and spatial resolutions. In this paper, we investigate aconception of a new pan-sharpening scheme using the pulse coupled neural network(PCNN) in the nonsubsampled shearlet transform (NSST) domain. This can be donebased on two main steps. In the first step, the input MS and Pan images are individuallydecomposed into multi-scaled and multi-directional coefficients by NSST.Second, the PCNN is applied to the low-frequency coefficients, which are mergedby a weighted firing energy fusion rule utilizing the PCNN firing times. The detailcoefficients with higher matching value are chosen to be the fused detail coefficients.Lastly, the pan-sharpened image is generated by the inverse NSST. WorldView-2,GeoEye-1, and QuickBird satellite datasets are employed in the experiments whichdemonstrate that the investigated scheme gained the ability in preserving both thehigh spatial details and high spectral characteristics simultaneously without involvingabundant computation time. In addition, various image quality metrics such asCC, RMSE, RASE, ERGAS, SAM, Q4, QNR, and SCC are adopted to assess thespectral and spatial qualities of the pan-sharpened image. The experimental resultsand performance analysis illustrated that our scheme improved performance efficiencyand achieved superiority over other conventional techniques.