Pan-Sharpening Based on Sparse Representation


AYAS S., TUNÇ GÖRMÜŞ E. , EKİNCİ M.

25th Signal Processing and Communications Applications Conference (SIU), Antalya, Türkiye, 15 - 18 Mayıs 2017 identifier identifier

Özet

Pan-sharpening in remote sensing aims to obtain a multispectral image with high spectral and spatial information by combining spectral information of a low resolution multispectral image and spatial information of a high resolution panchromatic image. In this paper, a pan-sharpening method based on sparse representation was proposed. Firstly, a dictionary was learned from the multispectral image patches. Then, sparse coefficients of low resolution multispectral image and high resolution panchromatic image were calculated. Thus, pan-sharpened multispectral image was obtained by using these sparse coefficients and dictionary. The IKONOS satellite image was used to test the proposed method. The quantitative and visual results demonstrate the effectiveness of the proposed method in pan-sharpening of the remote sensing images.