![]() | Only 14 pages are availabe for public view |
Abstract This thesis concerns with Pan-sharpening of satellite images by fusing a Panchromatic (PAN) image of high spatial resolution with Multi-Spectral (MS) images with high spectral resolution. The objective is to have satellite images with both high spatial and spectral resolution. Empirical Mode Decomposition (EMD) is used as it decomposes an image to a set of Intrinsic Mode Functions (IMFs) and a residual. This decomposition is utilized in the fusion process of images. High Pass Filtering (HPF) is used to isolate the high-frequency components ofthe PAN image, and hence EMD is performed on the high-frequency components and the MS images for a subsequent fusion process. This technique achieves better results than those obtained with the HPF fusion technique alone. A similar treatment is also presented, but on wavelet coefficients of both PAN and MS images using the Discrete Wavelet Transform (DWT). Better results have been achieved with treatment. To better enhance the obtained results from the quality metrics perspective, Modified Central Force Optimization (MCFO) is utilized in this thesis subject to certain constraints. The objective of the optimization process is to estimate the fusion coefficients that maximize the required quality metrics of the obtained images. The obtained results of optimization of conventional fusion schemes and one of the efficient signal decomposition techniques (EMD) prove the feasibility of the optimization process to enhance the fusion results. Keywords: DWT, EMD, HPF, image fusion, MCFO, optimization technique, Pansharpening, satellite images, spatial resolution, spectral resolution |