Proposing a Continuous Spectral-Spatial Sparse Representation Model for Denoising Hyperspectral Images

Due to its high capability in acquiring spectral and spatial information, hyperspectral imaging technology has gained significant attention remote sensing. However, practice, it is impossible avoid noise hyper images due camera artifacts the external environment during acquisition transmission process. The presence of these hinders detection subtle differences between different materials image. Therefore, crucial minimize as much possible before performing any analysis interpretation. Removing from a preprocessing step that enhances image quality for various applications, including object recognition classification. challenge arises when we need remove additive white mean-spherohomogeneous Gaussian given Previous research suggested thinning noise-free parts can be effective removing noise. This article aims implement method proposed using programming language. involves extracting intra-band structure inter-band correlation while displaying common tank learning dictionary. In continuous thin coding phase, extracted maintain same achieve spectrum continuity. contrast, used encode bands. Furthermore, joint dictionary training algorithm obtain simultaneously describes content ensures resulting preserves correlations noise-removal