Colour to Greyscale Pixels: Re-seeing Greyscale Archived Aerial Photographs and Declassified Satellite CORONA Images Based on Image Fusion Techniques

Abstract

To date, archived aerial photographs or recently declassified satellite CORONA images have been used to support archaeological research. However the use of such images is limited to the interpretation of greyscale images, which can be improved by image analysis processes including filters, edge detection techniques, etc. This article presents a methodology to improve the interpretation of these images by adding colour from recently acquired satellites or airborne sensors. The methodology followed in this article, is based on the use of pan-sharpening algorithms which are applied to a set of an archived greyscale image and a recently colour image taken over the same area. These two images are geometrically corrected while their histograms are matched. Then, the two images are merged based on different pan-sharpening techniques such as Brovey transformation, high pass filters, multiplicative transformation, principal component analysis and wavelet decomposition. The final result is a colour image which can be further studied for interpretation purposes. A coherence image is also calculated since differences due to time might be also observed in the images (e.g. land use changes). The results presented here from different cases studies of Cyprus indicate that the proposed methodology may be applied in any other greyscale archive dataset. From the results is was shown that Brovey and high pass filter transform seem to be able to support better the interpretation of such archive information.