الفهرس 
Chapter 1. IntroductionOnly 14 pages are availabe for public view
 |  | from | 151 |  |  |
| | | from | 151 | | |
Abstract
In recent years, many satellites were launched, which provide various types
of images about the earth surface. These images have different spectral and spatial
characteristics. Image fusion techniques are used extensively to combine different
images having complementary information into one single composite. The fused
image has rich information that will improve the performance of image analysis
algorithms. Pan-sharpening is a pixel level fusion technique used to increase the
spatial resolution of the multispectral (MS) image using spatial information from
the high resolution (HR) panchromatic (Pan) image while preserving the spectral
information in the MS image. Super-Resolution (SR) is a fusion process for
reconstructing an HR image from several low resolution (LR) images covering the
same region in the world. If the images are sub-pixel shifted, then each LR image
contains different information. SR extends the classical single-frame image
reconstruction/restoration methods by simultaneously utilizing information from
multiple observed images to achieve resolutions higher than those of the original
images.
The thesis firstly introduces the most popular existing Pan-sharpening
algorithms. The performance of these algorithms varies both spectrally and
spatially. The limitations of these algorithms are also analyzed.
In addition, a new integrated adaptive principle component analysis (APCA)
and High Pass Filtering (HPF) Pan-sharpening method is introduced and studied.
Simulation results have shown that the proposed fusion method gives a
compromise between the spectral and spatial quality, and at the same time has low
computational and memory complexity.
IV
The optimal set of Non-Subsampled Contourlet Transform (NSCT)
parameters (number of decomposition levels, number of orientations for different
levels, different decomposition filters) for Pan-sharpening of remote sensing
images with different spatial resolution ratios using hybrid APCA and NSCT
methods are investigated. Changing the number of orientations in the same level
of decomposition in the NSCT has a small effect on both the spectral and spatial
quality. The spectral and spatial quality of Pan-sharpened images mainly depends
on the number of decomposition levels. Performance comparison of Pansharpening
by the NSCT, the hybrid APCA/NSCT, the hybrid APCA/HPF, the
APCA, the HPF, the Gram–Schmidt (GS), the (Discrete Wavelet Transform )
DWT, the Stationary Wavelet Transform (SWT), hybrid APCA/DWT, and hybrid
APCA/SWT methods have been conducted. The results show that the Pansharpening
method with APCA/HPF has the best spectral and spatial quality, and
at the same time has low computational and memory complexity.
Finally, new efficient techniques for sharpening Misrsat-1 data using SR and
fusion methods are proposed. Due to the difference in spectral characteristics
between bands 1, 3, and the Pan, we proposed two different methods to enhance
the spatial resolution of Misrsat-1 data. The first method uses spatial-domain SR,
in which SR is performed on the high-pass details extracted from bands 1, 3, and
the Pan. The super-resolved high-pass details are used after that to enhance the
spatial resolution of the MS data using the HPF fusion method. The second
method depends on the interpolation of the high-frequency sub-band coefficients
of the multi-scale representations of bands 1, 3, and the Pan band and the additive
fusion method to add the high-frequency sub-band coefficients to different bands
of the MS image.