Hyperspectral Image Classification via a Novel Spectral–Spatial 3D ConvLSTM-CNN

In recent years, deep learning-based models have produced encouraging results for hyperspectral image (HSI) classification. Specifically, Convolutional Long Short-Term Memory (ConvLSTM) has shown good performance for learning valuable features and modeling long-term dependencies in spectral data. Ho...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Ghulam Farooque, Liang Xiao, Jingxiang Yang, Allah Bux Sargano
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	3D ConvLSTM hyperspectral image classification 3D CNN spectral-spatial feature extraction residual network deep learning Science Q
Acceso en línea:	https://doaj.org/article/adf4712f0c9e405fad3ea9075b2e76ae
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:adf4712f0c9e405fad3ea9075b2e76ae
record_format	dspace
spelling	oai:doaj.org-article:adf4712f0c9e405fad3ea9075b2e76ae2021-11-11T18:54:29ZHyperspectral Image Classification via a Novel Spectral–Spatial 3D ConvLSTM-CNN10.3390/rs132143482072-4292https://doaj.org/article/adf4712f0c9e405fad3ea9075b2e76ae2021-10-01T00:00:00Zhttps://www.mdpi.com/2072-4292/13/21/4348https://doaj.org/toc/2072-4292In recent years, deep learning-based models have produced encouraging results for hyperspectral image (HSI) classification. Specifically, Convolutional Long Short-Term Memory (ConvLSTM) has shown good performance for learning valuable features and modeling long-term dependencies in spectral data. However, it is less effective for learning spatial features, which is an integral part of hyperspectral images. Alternatively, convolutional neural networks (CNNs) can learn spatial features, but they possess limitations in handling long-term dependencies due to the local feature extraction in these networks. Considering these factors, this paper proposes an end-to-end Spectral-Spatial 3D ConvLSTM-CNN based Residual Network (SSCRN), which combines 3D ConvLSTM and 3D CNN for handling both spectral and spatial information, respectively. The contribution of the proposed network is twofold. Firstly, it addresses the long-term dependencies of spectral dimension using 3D ConvLSTM to capture the information related to various ground materials effectively. Secondly, it learns the discriminative spatial features using 3D CNN by employing the concept of the residual blocks to accelerate the training process and alleviate the overfitting. In addition, SSCRN uses batch normalization and dropout to regularize the network for smooth learning. The proposed framework is evaluated on three benchmark datasets widely used by the research community. The results confirm that SSCRN outperforms state-of-the-art methods with an overall accuracy of 99.17%, 99.67%, and 99.31% over Indian Pines, Salinas, and Pavia University datasets, respectively. Moreover, it is worth mentioning that these excellent results were achieved with comparatively fewer epochs, which also confirms the fast learning capabilities of the SSCRN.Ghulam FarooqueLiang XiaoJingxiang YangAllah Bux SarganoMDPI AGarticle3D ConvLSTMhyperspectral image classification3D CNNspectral-spatial feature extractionresidual networkdeep learningScienceQENRemote Sensing, Vol 13, Iss 4348, p 4348 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	3D ConvLSTM hyperspectral image classification 3D CNN spectral-spatial feature extraction residual network deep learning Science Q
spellingShingle	3D ConvLSTM hyperspectral image classification 3D CNN spectral-spatial feature extraction residual network deep learning Science Q Ghulam Farooque Liang Xiao Jingxiang Yang Allah Bux Sargano Hyperspectral Image Classification via a Novel Spectral–Spatial 3D ConvLSTM-CNN
description	In recent years, deep learning-based models have produced encouraging results for hyperspectral image (HSI) classification. Specifically, Convolutional Long Short-Term Memory (ConvLSTM) has shown good performance for learning valuable features and modeling long-term dependencies in spectral data. However, it is less effective for learning spatial features, which is an integral part of hyperspectral images. Alternatively, convolutional neural networks (CNNs) can learn spatial features, but they possess limitations in handling long-term dependencies due to the local feature extraction in these networks. Considering these factors, this paper proposes an end-to-end Spectral-Spatial 3D ConvLSTM-CNN based Residual Network (SSCRN), which combines 3D ConvLSTM and 3D CNN for handling both spectral and spatial information, respectively. The contribution of the proposed network is twofold. Firstly, it addresses the long-term dependencies of spectral dimension using 3D ConvLSTM to capture the information related to various ground materials effectively. Secondly, it learns the discriminative spatial features using 3D CNN by employing the concept of the residual blocks to accelerate the training process and alleviate the overfitting. In addition, SSCRN uses batch normalization and dropout to regularize the network for smooth learning. The proposed framework is evaluated on three benchmark datasets widely used by the research community. The results confirm that SSCRN outperforms state-of-the-art methods with an overall accuracy of 99.17%, 99.67%, and 99.31% over Indian Pines, Salinas, and Pavia University datasets, respectively. Moreover, it is worth mentioning that these excellent results were achieved with comparatively fewer epochs, which also confirms the fast learning capabilities of the SSCRN.
format	article
author	Ghulam Farooque Liang Xiao Jingxiang Yang Allah Bux Sargano
author_facet	Ghulam Farooque Liang Xiao Jingxiang Yang Allah Bux Sargano
author_sort	Ghulam Farooque
title	Hyperspectral Image Classification via a Novel Spectral–Spatial 3D ConvLSTM-CNN
title_short	Hyperspectral Image Classification via a Novel Spectral–Spatial 3D ConvLSTM-CNN
title_full	Hyperspectral Image Classification via a Novel Spectral–Spatial 3D ConvLSTM-CNN
title_fullStr	Hyperspectral Image Classification via a Novel Spectral–Spatial 3D ConvLSTM-CNN
title_full_unstemmed	Hyperspectral Image Classification via a Novel Spectral–Spatial 3D ConvLSTM-CNN
title_sort	hyperspectral image classification via a novel spectral–spatial 3d convlstm-cnn
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/adf4712f0c9e405fad3ea9075b2e76ae
work_keys_str_mv	AT ghulamfarooque hyperspectralimageclassificationviaanovelspectralspatial3dconvlstmcnn AT liangxiao hyperspectralimageclassificationviaanovelspectralspatial3dconvlstmcnn AT jingxiangyang hyperspectralimageclassificationviaanovelspectralspatial3dconvlstmcnn AT allahbuxsargano hyperspectralimageclassificationviaanovelspectralspatial3dconvlstmcnn
_version_	1718431665662984192

Hyperspectral Image Classification via a Novel Spectral–Spatial 3D ConvLSTM-CNN

Ejemplares similares