Pansharpening of Panchromatic Aerial Photographs by Combining Component Substitution and Multiresolution Analysis Fusion Algorithms

Abstract

The monochromatic nature of aerial photographs may affect the accuracy of photogrammetric measurements. In this study, multispectral (MS) IKONOS images with 3.2 m spatial resolution are used to sharpen panchromatic (PAN) archival aerial photographs. An urban area with a variety of land use/cover (LULC) classes over the northeast region of Cairo city, Egypt has been selected. The proposed approach can be performed through two major steps. First and after pre-processing, Gram-Schmidt transform (GST) and wavelet fusion techniques were used to panshrpen the original MS bands. Second, a fuzzy majority voting (FMV) - based approach was designed to combine the results from the GST and wavelet techniques to predict the final high-resolution results. The results were visually, qualitatively, and quantitatively analyzed. Compared with GST and wavelet standard fusion techniques, the improved statistical indices, as well as the improved classification potential, confirmed the credibility of the proposed approach.