On the use of the SOS metaheuristic algorithm in hybrid image fusion methods to achieve optimum spectral fidelity

Yilmaz Ç., Yilmaz V., GÜNGÖR O.

INTERNATIONAL JOURNAL OF REMOTE SENSING, vol.41, no.10, pp.3993-4021, 2020 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 41 Issue: 10
  • Publication Date: 2020
  • Doi Number: 10.1080/01431161.2019.1711244
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, PASCAL, Aerospace Database, Applied Science & Technology Source, Aqualine, Aquatic Science & Fisheries Abstracts (ASFA), Artic & Antarctic Regions, BIOSIS, CAB Abstracts, Communication Abstracts, Compendex, Computer & Applied Sciences, Environment Index, Geobase, INSPEC, Metadex, Pollution Abstracts, Public Affairs Index, Veterinary Science Database, DIALNET, Civil Engineering Abstracts
  • Page Numbers: pp.3993-4021
  • Karadeniz Technical University Affiliated: Yes


Image fusion aims to spatially enhance a low-resolution multispectral (MS) image by utilizing a high-resolution panchromatic (Pan) band. Various image fusion methodologies have been proposed with the aim to improve the spatial detail quality without deteriorating the colour content of the input MS image. Previous studies revealed the fact that there is no such thing as 'the best image fusion method' since all fusion methods cause either spectral distortion or spatial detail loss to some extent, which motivates the researchers to develop more advanced methods to keep the colour content while increasing the spatial detail quality. This study proposed to use the Symbiotic Organisms Search (SOS) metaheuristic algorithm in hybrid image fusion methods to achieve the optimum colour quality in the fused images. The SOS algorithm was used in two hybrid fusion approaches, one including the Intensity-Hue-Saturation (IHS) and Discrete Wavelet Transform (DWT) methods and the other one including the IHS and Discrete Wavelet Frame Transform (DWFT) methods. The results of the proposed methods were qualitatively and quantitatively compared in three test sites against those of eighteen widely-used image fusion methods. It was concluded that the proposed methods led to superior colour quality with both singlesensor and multisensor input images, regardless of the spatial resolution difference between the input images. The proposed methods were also found to be very successful in sharpening the images, despite the fact that their main purpose was to keep the colour content as much as possible.