Recent Trends in Satellite Image Pan-sharpening techniques
Résumé
Nowadays, there are several satellites bearing sensors to provide images from different spectral bands at the same time, but with various resolutions and at different frequencies. In practice, there is an image captured over a wide spectral band which is known as panchromatic (PAN) image. Besides this image there are at least three multispectral (MS) images captured over narrower spectral bands including the visible red (R), green (G) and blue (B) bands. Commonly, multispectral images have lower spatial resolution than that of panchromatic image. Though, today's sensors can produce multispectral images with higher spatial resolution than before, it is still desirable to enhance their resolution in order to approach that of the PAN image. The process of generating high resolution color images from MS and PAN images is known as pan-sharpening. This process fuses the low-resolution MS images with the higher-resolution PAN image so that the resulting images corresponding to MS images are of the same resolution as the PAN image. This helps to obtain color images with higher spatial resolution for better visualization of remote sensing color images. This paper presents a review of pan-sharpening methods and then analyzes the trends of the algorithms. Through this review, we will analyze the impact of pan-sharpening on remote sensing images processing, and particularly, we will observe the income of pan-sharpening in vegetation detection. Based on our previous work, a strategy is implemented to evaluate the performance of pan-sharpened images.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...