Pansharpening is considered as an imperative process for various remote sensing applications viz. crop monitoring, hazard monitoring, object detection and classification etc. The Pansharpening technique combines panchromatic and multispectral pictures to create a high resolution multispectral image. In this paper, the pansharpening approach and a variational optimization model are discussed. As an ill-posed inverse issue, a cost function is proposed, with three prior components, two of which are data-fidelity terms generated from the relationship between the source and output images. The third term is integrated to regularize the formulated inverse model. The eminent solver, alternating direction method of multipliers in conjunction with iterative minimization mechanism is employed to obtain the comprehensive minimum of the proposed convex cost function. The minimized solution is the required pansharpened image. The effectiveness of the suggested strategy is assessed using three different datasets and four recognized indicators. The results, both objective and subjective, show the effectiveness of the variational optimization pansharpening (VOPS) model. The merged image has greatly improved spectral and spatial properties.
Pansharpening of Multispectral Images Through the Inverse Problem Model with Non-convex Sparse Regularization
Lect. Notes Electrical Eng.
International Conference on Robotics, Control, Automation and Artificial Intelligence ; 2022 November 24, 2022 - November 26, 2022
2023-11-18
13 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Fast Weighted Least Squares Based Pansharpening
IEEE | 2019
|Convex Image Denoising via Non-convex Regularization with Parameter Selection
British Library Online Contents | 2016
|