Spatio-Spectral Structure Tensor Total Variation for Hyperspectral Image Denoising and Destriping

This article proposes a novel regularization method, named <italic>spatio-spectral structure tensor total variation</italic> (<inline-formula><tex-math notation="LaTeX">${\text{S}_{3} \text{TTV}}$</tex-math></inline-formula>), for denoising and destripin...

Full description

Saved in:
Bibliographic Details
Main Authors: Shingo Takemoto, Kazuki Naganuma, Shunsuke Ono
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11072325/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850034631132315648
author Shingo Takemoto
Kazuki Naganuma
Shunsuke Ono
author_facet Shingo Takemoto
Kazuki Naganuma
Shunsuke Ono
author_sort Shingo Takemoto
collection DOAJ
description This article proposes a novel regularization method, named <italic>spatio-spectral structure tensor total variation</italic> (<inline-formula><tex-math notation="LaTeX">${\text{S}_{3} \text{TTV}}$</tex-math></inline-formula>), for denoising and destriping hyperspectral (HS) images. HS images are inevitably contaminated by various types of noise during acquisition process due to the measurement equipment and the environment. For HS image denoising and destriping tasks, spatio-spectral total variation (SSTV) is widely known as a powerful regularization approach that models the spatio-spectral piecewise smoothness. However, since SSTV refers only to the local differences of pixels/bands, edges and textures that extend beyond adjacent pixels are not preserved during denoising process. To address this problem, we newly introduce <inline-formula><tex-math notation="LaTeX">${\text{S}_{3} \text{TTV}}$</tex-math></inline-formula>, which is designed to preserve two essential physical characteristics of HS images: semilocal spatial structures and spectral correlation across all bands. Specifically, we define <inline-formula><tex-math notation="LaTeX">${\text{S}_{3} \text{TTV}}$</tex-math></inline-formula> as the sum of the nuclear norms of spatio-spectral structure tensors, which are matrices formed by arranging second-order spatio-spectral difference vectors within semilocal areas. Furthermore, we formulate the HS image denoising and destriping problem as a constrained convex optimization problem involving <inline-formula><tex-math notation="LaTeX">${\text{S}_{3} \text{TTV}}$</tex-math></inline-formula> and develop an algorithm based on a preconditioned primal-dual splitting method to solve this problem efficiently. Finally, we demonstrate the effectiveness of <inline-formula><tex-math notation="LaTeX">${\text{S}_{3} \text{TTV}}$</tex-math></inline-formula> by comparing it with existing methods, including state-of-the-art ones through denoising and destriping experiments.
format Article
id doaj-art-a4f691ac96ac4ffcb5abbaf421b24f83
institution DOAJ
issn 1939-1404
2151-1535
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
spelling doaj-art-a4f691ac96ac4ffcb5abbaf421b24f832025-08-20T02:57:45ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-0118191571917510.1109/JSTARS.2025.358677911072325Spatio-Spectral Structure Tensor Total Variation for Hyperspectral Image Denoising and DestripingShingo Takemoto0https://orcid.org/0000-0002-8075-8983Kazuki Naganuma1https://orcid.org/0000-0002-7180-3017Shunsuke Ono2https://orcid.org/0000-0001-7890-5131Department of Computer Science, Institute of Science Tokyo, Yokohama, JapanInstitute of Engineering of Tokyo University of Agriculture and Technology, Tokyo, JapanDepartment of Computer Science, Institute of Science Tokyo, Yokohama, JapanThis article proposes a novel regularization method, named <italic>spatio-spectral structure tensor total variation</italic> (<inline-formula><tex-math notation="LaTeX">${\text{S}_{3} \text{TTV}}$</tex-math></inline-formula>), for denoising and destriping hyperspectral (HS) images. HS images are inevitably contaminated by various types of noise during acquisition process due to the measurement equipment and the environment. For HS image denoising and destriping tasks, spatio-spectral total variation (SSTV) is widely known as a powerful regularization approach that models the spatio-spectral piecewise smoothness. However, since SSTV refers only to the local differences of pixels/bands, edges and textures that extend beyond adjacent pixels are not preserved during denoising process. To address this problem, we newly introduce <inline-formula><tex-math notation="LaTeX">${\text{S}_{3} \text{TTV}}$</tex-math></inline-formula>, which is designed to preserve two essential physical characteristics of HS images: semilocal spatial structures and spectral correlation across all bands. Specifically, we define <inline-formula><tex-math notation="LaTeX">${\text{S}_{3} \text{TTV}}$</tex-math></inline-formula> as the sum of the nuclear norms of spatio-spectral structure tensors, which are matrices formed by arranging second-order spatio-spectral difference vectors within semilocal areas. Furthermore, we formulate the HS image denoising and destriping problem as a constrained convex optimization problem involving <inline-formula><tex-math notation="LaTeX">${\text{S}_{3} \text{TTV}}$</tex-math></inline-formula> and develop an algorithm based on a preconditioned primal-dual splitting method to solve this problem efficiently. Finally, we demonstrate the effectiveness of <inline-formula><tex-math notation="LaTeX">${\text{S}_{3} \text{TTV}}$</tex-math></inline-formula> by comparing it with existing methods, including state-of-the-art ones through denoising and destriping experiments.https://ieeexplore.ieee.org/document/11072325/Denoisingdestripinghyperspectral imagespatio-spectral regularizationstructure tensortotal variation
spellingShingle Shingo Takemoto
Kazuki Naganuma
Shunsuke Ono
Spatio-Spectral Structure Tensor Total Variation for Hyperspectral Image Denoising and Destriping
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Denoising
destriping
hyperspectral image
spatio-spectral regularization
structure tensor
total variation
title Spatio-Spectral Structure Tensor Total Variation for Hyperspectral Image Denoising and Destriping
title_full Spatio-Spectral Structure Tensor Total Variation for Hyperspectral Image Denoising and Destriping
title_fullStr Spatio-Spectral Structure Tensor Total Variation for Hyperspectral Image Denoising and Destriping
title_full_unstemmed Spatio-Spectral Structure Tensor Total Variation for Hyperspectral Image Denoising and Destriping
title_short Spatio-Spectral Structure Tensor Total Variation for Hyperspectral Image Denoising and Destriping
title_sort spatio spectral structure tensor total variation for hyperspectral image denoising and destriping
topic Denoising
destriping
hyperspectral image
spatio-spectral regularization
structure tensor
total variation
url https://ieeexplore.ieee.org/document/11072325/
work_keys_str_mv AT shingotakemoto spatiospectralstructuretensortotalvariationforhyperspectralimagedenoisinganddestriping
AT kazukinaganuma spatiospectralstructuretensortotalvariationforhyperspectralimagedenoisinganddestriping
AT shunsukeono spatiospectralstructuretensortotalvariationforhyperspectralimagedenoisinganddestriping