Research and evaluation of Beijing-1 image fusion based on Imagesharp algorithm

Abstract

Beijing-1 small satellite carries a 4m panchromatic sensor and a 32m multi-spectral sensor, with both the features of the high resolution of SPOT and the multi-spectrum of Landsat. However, its resolution and the wavelength range of the corresponding band are significantly different from the existing satellites. Meanwhile, the difference of the wavelength ranges between the panchromatic image and the multi-spectral image is rather large, especially in the NIR band where the difference reaches 80%. Thus, low correlation means that the traditional image fusion methods are not ideal to date. Aimed at the spectral features of Beijing-1 small satellite high-resolution imagery and multi-spectral imagery, the paper proposes an image fusion method based on Imagesharp. First, linear functions are used to fit the low-resolution image into the high-resolution image, and then the fitted high-resolution multi-spectral image and the high-resolution panchromatic image are fused, which can improve the spatial resolution and keep the original multi-spectral information well at the same time. At last, in comparison to the traditional HIS and Brovey fusion method, Imagesharp method can maintain the color saturation and the spectral information better. A large amount of experiments prove that Imagesharp algorithm is suitable for the Beijing-1 image fusion and it is already specially applied in the Beijing-1 small satellite data deep processing software.