A Flexible Image Fusion Algorithm Applied on Fusion Between Optical and Thermal Data

Abstract

Classic image fusion algorithms fail to apply to thermal images because they require at least three bands as input but thermal images have only one band. Recent studies on fusing visible and thermal images restrict to close-range photographic images or require abundant image samples. This paper presents an image fusion algorithm that is flexible with different spectrum sensors, as well as the number of bands. This algorithm transfers the image into the frequency domain, where the spatial information from different sensors can be separated, extracted, and then synthesized. The algorithm was implemented in image fusion between an aerial photo and thermal image, as well as the space-borne thermal and visible band. Both cases realized spatial enhancement and spectral preservation. © 2022 IEEE.