Hyperspectral Image Classification Based on Multilevel Joint Feature Extraction Network

Over the past few years, convolutional neural network (CNN) has been broadly adopted in remote sensing (RS) imagery processing areas due to its impressive capabilities in feature extraction. Nevertheless, it is still a challenge for CNN-based hyperspectral image (HSI) classification methods to extra...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Xiaochen Lu, Dezheng Yang, Fengde Jia, Yunlong Yang, Lei Zhang
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2021
Materias:	Attention details convolutional neural network (CNN) hyperspectral image (HSI) image classification multilevel feature extraction (MFE) Ocean engineering TC1501-1800 Geophysics. Cosmic physics QC801-809
Acceso en línea:	https://doaj.org/article/e92dd3f77f3e4f1aba65b48c4ec7ec84
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:e92dd3f77f3e4f1aba65b48c4ec7ec84
record_format	dspace
spelling	oai:doaj.org-article:e92dd3f77f3e4f1aba65b48c4ec7ec842021-11-18T00:00:23ZHyperspectral Image Classification Based on Multilevel Joint Feature Extraction Network2151-153510.1109/JSTARS.2021.3123371https://doaj.org/article/e92dd3f77f3e4f1aba65b48c4ec7ec842021-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/9591315/https://doaj.org/toc/2151-1535Over the past few years, convolutional neural network (CNN) has been broadly adopted in remote sensing (RS) imagery processing areas due to its impressive capabilities in feature extraction. Nevertheless, it is still a challenge for CNN-based hyperspectral image (HSI) classification methods to extract more effective spectral-spatial features considering all spectral bands. Driven by this issue, we propose a novel approach to cope with the HSI classification task, referring to the multilevel joint feature extraction network. The proposed network makes full use of the information on each channel of HSI and transforms it into valid channel-wised spatial features through a designed convolution process. Moreover, these feature maps form global attention details to guide the extraction of spectral-spatial features, which are taken to the next level for further feature mining. Then, the features obtained at different levels are integrated for ground object classification. In contrast with several state-of-the-art HSI classification methods on four public datasets, experimental results demonstrate the effectiveness and remarkable feature extraction capability of our proposed approach.Xiaochen LuDezheng YangFengde JiaYunlong YangLei ZhangIEEEarticleAttention detailsconvolutional neural network (CNN)hyperspectral image (HSI)image classificationmultilevel feature extraction (MFE)Ocean engineeringTC1501-1800Geophysics. Cosmic physicsQC801-809ENIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, Vol 14, Pp 10977-10989 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Attention details convolutional neural network (CNN) hyperspectral image (HSI) image classification multilevel feature extraction (MFE) Ocean engineering TC1501-1800 Geophysics. Cosmic physics QC801-809
spellingShingle	Attention details convolutional neural network (CNN) hyperspectral image (HSI) image classification multilevel feature extraction (MFE) Ocean engineering TC1501-1800 Geophysics. Cosmic physics QC801-809 Xiaochen Lu Dezheng Yang Fengde Jia Yunlong Yang Lei Zhang Hyperspectral Image Classification Based on Multilevel Joint Feature Extraction Network
description	Over the past few years, convolutional neural network (CNN) has been broadly adopted in remote sensing (RS) imagery processing areas due to its impressive capabilities in feature extraction. Nevertheless, it is still a challenge for CNN-based hyperspectral image (HSI) classification methods to extract more effective spectral-spatial features considering all spectral bands. Driven by this issue, we propose a novel approach to cope with the HSI classification task, referring to the multilevel joint feature extraction network. The proposed network makes full use of the information on each channel of HSI and transforms it into valid channel-wised spatial features through a designed convolution process. Moreover, these feature maps form global attention details to guide the extraction of spectral-spatial features, which are taken to the next level for further feature mining. Then, the features obtained at different levels are integrated for ground object classification. In contrast with several state-of-the-art HSI classification methods on four public datasets, experimental results demonstrate the effectiveness and remarkable feature extraction capability of our proposed approach.
format	article
author	Xiaochen Lu Dezheng Yang Fengde Jia Yunlong Yang Lei Zhang
author_facet	Xiaochen Lu Dezheng Yang Fengde Jia Yunlong Yang Lei Zhang
author_sort	Xiaochen Lu
title	Hyperspectral Image Classification Based on Multilevel Joint Feature Extraction Network
title_short	Hyperspectral Image Classification Based on Multilevel Joint Feature Extraction Network
title_full	Hyperspectral Image Classification Based on Multilevel Joint Feature Extraction Network
title_fullStr	Hyperspectral Image Classification Based on Multilevel Joint Feature Extraction Network
title_full_unstemmed	Hyperspectral Image Classification Based on Multilevel Joint Feature Extraction Network
title_sort	hyperspectral image classification based on multilevel joint feature extraction network
publisher	IEEE
publishDate	2021
url	https://doaj.org/article/e92dd3f77f3e4f1aba65b48c4ec7ec84
work_keys_str_mv	AT xiaochenlu hyperspectralimageclassificationbasedonmultileveljointfeatureextractionnetwork AT dezhengyang hyperspectralimageclassificationbasedonmultileveljointfeatureextractionnetwork AT fengdejia hyperspectralimageclassificationbasedonmultileveljointfeatureextractionnetwork AT yunlongyang hyperspectralimageclassificationbasedonmultileveljointfeatureextractionnetwork AT leizhang hyperspectralimageclassificationbasedonmultileveljointfeatureextractionnetwork
_version_	1718425258032103424

Hyperspectral Image Classification Based on Multilevel Joint Feature Extraction Network

Ejemplares similares