Superpixel-Based Singular Spectrum Analysis for Effective Spatial-Spectral Feature Extraction

In the processing of remotely sensed data, classification may be preceded by feature extraction, which helps in making the most informative parts of the data emerge. Effective feature extraction may boost the efficiency and accuracy of the following classification, and hence various methods have bee...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Subhashree Subudhi, Ramnarayan Patro , Pradyut Kumar Biswal, Fabio Dell’Acqua
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	hyperspectral image superpixel segmentation evaluation 2D-singular spectrum analysis (2D-SSA) feature extraction Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999
Acceso en línea:	https://doaj.org/article/2f877d43845a449c90418d672844c38d
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:2f877d43845a449c90418d672844c38d
record_format	dspace
spelling	oai:doaj.org-article:2f877d43845a449c90418d672844c38d2021-11-25T16:39:41ZSuperpixel-Based Singular Spectrum Analysis for Effective Spatial-Spectral Feature Extraction10.3390/app1122108762076-3417https://doaj.org/article/2f877d43845a449c90418d672844c38d2021-11-01T00:00:00Zhttps://www.mdpi.com/2076-3417/11/22/10876https://doaj.org/toc/2076-3417In the processing of remotely sensed data, classification may be preceded by feature extraction, which helps in making the most informative parts of the data emerge. Effective feature extraction may boost the efficiency and accuracy of the following classification, and hence various methods have been proposed to perform it. Recently, Singular Spectrum Analysis (SSA) and its 2-D variation (2D-SSA) have emerged as popular, cutting-edge technologies for effective feature extraction in Hyperspectral Images (HSI). Using 2D-SSA, each band image of an HSI is initially decomposed into various components, and then the image is reconstructed using the most significant eigen-tuples relative to their eigen-values, which represent strong spatial features for the classification task. However, instead of performing reconstruction on the whole image, it may be more effective to apply reconstruction to object-specific spatial regions, which is the proposed objective of this research. As an HSI may cover a large area, multiple objects are generally present within a single scene. Hence, spatial information can be highlighted accurately by specializing the reconstruction based on the local context. The local context may be defined by the so-called superpixels, i.e., finite sets of pixels that constitute a homogeneous set. Each superpixel may undergo tailored reconstruction, with a process expected to perform better than non-spatially-adaptive approaches. In this paper, a Superpixel-based SSA (SP-SSA) method is proposed where the image is first segmented into multiple regions using a superpixel segmentation approach. Next, each segment is individually reconstructed using 2D-SSA. In doing so, the spatial contextual information is preserved, leading to better classifier performance. The performance of the reconstructed features is evaluated using an SVM classifier. Experiments on four popular benchmark datasets reveal that, in terms of the classification accuracy, the proposed approach overperforms the standard SSA technique and various common spatio-spectral classification methods.Subhashree SubudhiRamnarayan Patro Pradyut Kumar BiswalFabio Dell’AcquaMDPI AGarticlehyperspectral imagesuperpixel segmentationevaluation2D-singular spectrum analysis (2D-SSA)feature extractionTechnologyTEngineering (General). Civil engineering (General)TA1-2040Biology (General)QH301-705.5PhysicsQC1-999ChemistryQD1-999ENApplied Sciences, Vol 11, Iss 10876, p 10876 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	hyperspectral image superpixel segmentation evaluation 2D-singular spectrum analysis (2D-SSA) feature extraction Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999
spellingShingle	hyperspectral image superpixel segmentation evaluation 2D-singular spectrum analysis (2D-SSA) feature extraction Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999 Subhashree Subudhi Ramnarayan Patro Pradyut Kumar Biswal Fabio Dell’Acqua Superpixel-Based Singular Spectrum Analysis for Effective Spatial-Spectral Feature Extraction
description	In the processing of remotely sensed data, classification may be preceded by feature extraction, which helps in making the most informative parts of the data emerge. Effective feature extraction may boost the efficiency and accuracy of the following classification, and hence various methods have been proposed to perform it. Recently, Singular Spectrum Analysis (SSA) and its 2-D variation (2D-SSA) have emerged as popular, cutting-edge technologies for effective feature extraction in Hyperspectral Images (HSI). Using 2D-SSA, each band image of an HSI is initially decomposed into various components, and then the image is reconstructed using the most significant eigen-tuples relative to their eigen-values, which represent strong spatial features for the classification task. However, instead of performing reconstruction on the whole image, it may be more effective to apply reconstruction to object-specific spatial regions, which is the proposed objective of this research. As an HSI may cover a large area, multiple objects are generally present within a single scene. Hence, spatial information can be highlighted accurately by specializing the reconstruction based on the local context. The local context may be defined by the so-called superpixels, i.e., finite sets of pixels that constitute a homogeneous set. Each superpixel may undergo tailored reconstruction, with a process expected to perform better than non-spatially-adaptive approaches. In this paper, a Superpixel-based SSA (SP-SSA) method is proposed where the image is first segmented into multiple regions using a superpixel segmentation approach. Next, each segment is individually reconstructed using 2D-SSA. In doing so, the spatial contextual information is preserved, leading to better classifier performance. The performance of the reconstructed features is evaluated using an SVM classifier. Experiments on four popular benchmark datasets reveal that, in terms of the classification accuracy, the proposed approach overperforms the standard SSA technique and various common spatio-spectral classification methods.
format	article
author	Subhashree Subudhi Ramnarayan Patro Pradyut Kumar Biswal Fabio Dell’Acqua
author_facet	Subhashree Subudhi Ramnarayan Patro Pradyut Kumar Biswal Fabio Dell’Acqua
author_sort	Subhashree Subudhi
title	Superpixel-Based Singular Spectrum Analysis for Effective Spatial-Spectral Feature Extraction
title_short	Superpixel-Based Singular Spectrum Analysis for Effective Spatial-Spectral Feature Extraction
title_full	Superpixel-Based Singular Spectrum Analysis for Effective Spatial-Spectral Feature Extraction
title_fullStr	Superpixel-Based Singular Spectrum Analysis for Effective Spatial-Spectral Feature Extraction
title_full_unstemmed	Superpixel-Based Singular Spectrum Analysis for Effective Spatial-Spectral Feature Extraction
title_sort	superpixel-based singular spectrum analysis for effective spatial-spectral feature extraction
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/2f877d43845a449c90418d672844c38d
work_keys_str_mv	AT subhashreesubudhi superpixelbasedsingularspectrumanalysisforeffectivespatialspectralfeatureextraction AT ramnarayanpatro superpixelbasedsingularspectrumanalysisforeffectivespatialspectralfeatureextraction AT pradyutkumarbiswal superpixelbasedsingularspectrumanalysisforeffectivespatialspectralfeatureextraction AT fabiodellacqua superpixelbasedsingularspectrumanalysisforeffectivespatialspectralfeatureextraction
_version_	1718413067863195648

Superpixel-Based Singular Spectrum Analysis for Effective Spatial-Spectral Feature Extraction

Ejemplares similares