Dimensionality reduction method based on spatial-spectral preservation and minimum noise fraction for hyperspectral images

Hyperspectral images contain rich spatial distribution and spectral information of land features, but they also introduce high information redundancy and computational complexity. This paper proposes dimensionality reduction methods that integrate spatial-spectral preservation and minimum noise frac...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Zhou Bing, Deng Lei, Ying Jiaju, Wang Qianghui, Cheng Yue
Format: Artikel
Sprache:Englisch
Veröffentlicht: EDP Sciences 2025-01-01
Schriftenreihe:Journal of the European Optical Society-Rapid Publications
Schlagworte:
Online-Zugang:https://jeos.edpsciences.org/articles/jeos/full_html/2025/02/jeos20250029/jeos20250029.html
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1839624729500057600
author Zhou Bing
Deng Lei
Ying Jiaju
Wang Qianghui
Cheng Yue
author_facet Zhou Bing
Deng Lei
Ying Jiaju
Wang Qianghui
Cheng Yue
author_sort Zhou Bing
collection DOAJ
description Hyperspectral images contain rich spatial distribution and spectral information of land features, but they also introduce high information redundancy and computational complexity. This paper proposes dimensionality reduction methods that integrate spatial-spectral preservation and minimum noise fraction (MNF) to better analyze and utilize the spatial and spectral information in hyperspectral images. While performing the minimum noise separation transformation, the proposed method aims to preserve the spatial structure of the image as much as possible, maximizing both the signal-to-noise ratio and the spatial structure similarity of the image. The component selection strategy involves grouping components and calculating the average change in the relative position of all pixels in the feature space. The component group that most closely matches the spectral relative position before transformation is selected as the final dimensionality reduction result. Experimental results demonstrate that the proposed method is highly sensitive to noise estimation and requires a relatively accurate noise covariance matrix. The method effectively preserves spatial information, with negligible impact on the accuracy of object detection methods, and outperforms other comparative approaches. It ensures the effectiveness of downstream object detection tasks while significantly reducing computational time. The code of the proposed method is available at https://github.com/aosilu/spatial-spectral-preservation-MNF.
format Article
id doaj-art-e7965dae4b2b4e90b9398cc7e2fe6597
institution Matheson Library
issn 1990-2573
language English
publishDate 2025-01-01
publisher EDP Sciences
record_format Article
series Journal of the European Optical Society-Rapid Publications
spelling doaj-art-e7965dae4b2b4e90b9398cc7e2fe65972025-07-18T08:22:42ZengEDP SciencesJournal of the European Optical Society-Rapid Publications1990-25732025-01-012123110.1051/jeos/2025029jeos20250029Dimensionality reduction method based on spatial-spectral preservation and minimum noise fraction for hyperspectral imagesZhou Bing0Deng Lei1Ying Jiaju2Wang Qianghui3Cheng Yue4Shijiazhuang Campus, Army Engineering University of PLAShijiazhuang Campus, Army Engineering University of PLAShijiazhuang Campus, Army Engineering University of PLAState Key Laboratory of Complex Electromagnetic Environment Effects on Electronics and Information SystemShijiazhuang Campus, Army Engineering University of PLAHyperspectral images contain rich spatial distribution and spectral information of land features, but they also introduce high information redundancy and computational complexity. This paper proposes dimensionality reduction methods that integrate spatial-spectral preservation and minimum noise fraction (MNF) to better analyze and utilize the spatial and spectral information in hyperspectral images. While performing the minimum noise separation transformation, the proposed method aims to preserve the spatial structure of the image as much as possible, maximizing both the signal-to-noise ratio and the spatial structure similarity of the image. The component selection strategy involves grouping components and calculating the average change in the relative position of all pixels in the feature space. The component group that most closely matches the spectral relative position before transformation is selected as the final dimensionality reduction result. Experimental results demonstrate that the proposed method is highly sensitive to noise estimation and requires a relatively accurate noise covariance matrix. The method effectively preserves spatial information, with negligible impact on the accuracy of object detection methods, and outperforms other comparative approaches. It ensures the effectiveness of downstream object detection tasks while significantly reducing computational time. The code of the proposed method is available at https://github.com/aosilu/spatial-spectral-preservation-MNF.https://jeos.edpsciences.org/articles/jeos/full_html/2025/02/jeos20250029/jeos20250029.htmlhyperspectral imagedimensionality reductionminimum noise fraction (mnf)spatial-spectral preservation
spellingShingle Zhou Bing
Deng Lei
Ying Jiaju
Wang Qianghui
Cheng Yue
Dimensionality reduction method based on spatial-spectral preservation and minimum noise fraction for hyperspectral images
Journal of the European Optical Society-Rapid Publications
hyperspectral image
dimensionality reduction
minimum noise fraction (mnf)
spatial-spectral preservation
title Dimensionality reduction method based on spatial-spectral preservation and minimum noise fraction for hyperspectral images
title_full Dimensionality reduction method based on spatial-spectral preservation and minimum noise fraction for hyperspectral images
title_fullStr Dimensionality reduction method based on spatial-spectral preservation and minimum noise fraction for hyperspectral images
title_full_unstemmed Dimensionality reduction method based on spatial-spectral preservation and minimum noise fraction for hyperspectral images
title_short Dimensionality reduction method based on spatial-spectral preservation and minimum noise fraction for hyperspectral images
title_sort dimensionality reduction method based on spatial spectral preservation and minimum noise fraction for hyperspectral images
topic hyperspectral image
dimensionality reduction
minimum noise fraction (mnf)
spatial-spectral preservation
url https://jeos.edpsciences.org/articles/jeos/full_html/2025/02/jeos20250029/jeos20250029.html
work_keys_str_mv AT zhoubing dimensionalityreductionmethodbasedonspatialspectralpreservationandminimumnoisefractionforhyperspectralimages
AT denglei dimensionalityreductionmethodbasedonspatialspectralpreservationandminimumnoisefractionforhyperspectralimages
AT yingjiaju dimensionalityreductionmethodbasedonspatialspectralpreservationandminimumnoisefractionforhyperspectralimages
AT wangqianghui dimensionalityreductionmethodbasedonspatialspectralpreservationandminimumnoisefractionforhyperspectralimages
AT chengyue dimensionalityreductionmethodbasedonspatialspectralpreservationandminimumnoisefractionforhyperspectralimages