Abstract
Pansharpening algorithms are designed to enhance the spatial resolution of multispectral images using panchromatic images with high spatial resolutions. Panchromatic and multispectral images acquired from very high resolution (VHR) satellite sensors used as input data in the pansharpening process are characterized by spatial dissimilarities due to differences in their spectral/spatial characteristics and time lags between panchromatic and multispectral sensors. In this manuscript, a new pansharpening framework is proposed to improve the spatial clarity of VHR satellite imagery. This algorithm aims to remove the spatial dissimilarity between panchromatic and multispectral images using guided filtering (GF) and to generate the optimal local injection gains for pansharpening. First, we generate optimal multispectral images with spatial characteristics similar to those of panchromatic images using GF. Then, multiresolution analysis (MRA)-based pansharpening is applied using normalized difference vegetation index (NDVI)-based optimal injection gains and spatial details obtained through GF. The algorithm is applied to Korea multipurpose satellite (KOMPSAT)-3/3A satellite sensor data, and the experimental results show that the pansharpened images obtained with the proposed algorithm exhibit a superior spatial quality and preserve spectral information better than those based on existing algorithms.
Highlights
Very high resolution (VHR) satellite sensors, such as WorldView-3, Pléiades, and the Korea multipurpose satellite (KOMPSAT)-3/3A, provide panchromatic images with high spatial resolutions and multispectral images with low spatial resolutions
General pansharpening algorithms have been classified into component substitution (CS)-based and multiresolution analysis (MRA)-based methods depending on how the spatial details are generated [2]
GFNDVI denotes the proposed guided filtering (GF)-based pansharpening algorithm using local injection gains based on the normalized difference vegetation index (NDVI)
Summary
Very high resolution (VHR) satellite sensors, such as WorldView-3, Pléiades, and the Korea multipurpose satellite (KOMPSAT)-3/3A, provide panchromatic images with high spatial resolutions and multispectral images with low spatial resolutions. CS-based algorithms generate pansharpened images by adding spatial details based on high-frequency information from panchromatic images with a high spatial resolution and synthetic intensity images with a low spatial resolution [8,9,10]. The generalized intensity-hue-saturation (GIHS), Gram–Schmidt (GS), GS adaptive (GSA), and band-dependent spatial detail (BDSD) methods are representative CS-based pansharpening techniques [2,11,12] Hybrid algorithms, such as partial replacement adaptive component substitution (PRACS) and generalized BDSD algorithms, have been developed in addition to various CS algorithms using global and local injection gains [13,14,15,16]. Most researchers have developed various pansharpening algorithms based on either CS or MRA aimed at generating multispectral images with a spatial resolution similar to that of a panchromatic image while preserving the spectral information of the former [2,18]
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
