Traditional deep learning methods extract only deep abstract information from hyperspectral images while failing to fully reveal the local geometric structure relationship between samples. This limits the improvement of classification performance. To address this problem, the present study proposes a new feature extraction network called a deep manifold reconstruction belief network. First, deep abstract features are extracted based on the deep belief network to enhance the identification ability of abstract features. Then, intraclass and interclass graphs are constructed based on the neighborhood points of sample data and reconstruction points of similar neighbors in each neighborhood under the graph embedding framework. Under this framework, intraclass neighbors and their reconstruction points are compressed. By contrast, interclass neighbors and their reconstruction points are separated in low-dimensional space to improve the separability of different types of data and the accuracy of feature classification. Deep discriminant feature extraction is then realized based on the reconstructed points. Experimental results on the KSC and MUUFL Gulfport hyperspectral datasets showed that the overall classification accuracy of the proposed algorithm was 94.71% and 86.38%, respectively. Compared with other algorithms, the proposed algorithm effectively improves the ability of land cover classification and is more conducive to practical applications.
DING G SH , QIAO Y L , YI W N , et al . Improved SIFT feature extraction and matching technology based on hyperspectral image [J]. Opt. Precision Eng. , 2020 , 28 ( 4 ): 954 - 962 . (in Chinese)
DING G S , QIAO Y L , YI W N , et al .. Improved SIFT feature extraction and matching technology based on hyperspectral image [J]. Optics and Precision Engineering , 2020 , 28 ( 04 ): 954 - 962 .
WU F Y , WANG X , DING J W , et al . Improved cascade forest deep learning model for hyperspectral imagery classification [J]. Journal of Remote Sensing , 2020 , 24 ( 4 ): 439 - 453 . (in Chinese)
YAN J W , CHEN H D , LIU L . Overview of hyperspectral image classification [J]. Opt. Precision Eng. , 2019 , 27 ( 3 ): 680 - 693 . (in Chinese)
CAI Y M , LIU X B , CAI Z H . BS-nets: an end-to-end framework for band selection of hyperspectral image [J]. IEEE Transactions on Geoscience and Remote Sensing , 2020 , 58 ( 3 ): 1969 - 1984 .
HABERMANN M , FREMONT V , SHIGUEMORI E H . Supervised band selection in hyperspectral images using single-layer neural networks [J]. International Journal of Remote Sensing , 2019 , 40 ( 10 ): 3900 - 3926 .
LUO F L , DU B , ZHANG L P , et al . Feature learning using spatial-spectral hypergraph discriminant analysis for hyperspectral image [J]. IEEE Transactions on Cybernetics , 2019 , 49 ( 7 ): 2406 - 2419 .
BONIFAZI G , CAPOBIANCO G , SERRANTI S . Asbestos containing materials detection and classification by the use of hyperspectral imaging [J]. Journal of Hazardous Materials , 2018 , 344 : 981 - 993 .
WANG H , FAN Y Y , FANG B F , et al . Generalized linear discriminant analysis based on euclidean norm for gait recognition [J]. International Journal of Machine Learning and Cybernetics , 2018 , 9 ( 4 ): 569 - 576 .
DENG Y J , LI H C , PAN L , et al . Modified tensor locality preserving projection for dimensionality reduction of hyperspectral images [J]. IEEE Geoscience and Remote Sensing Letters , 2018 , 15 ( 2 ): 277 - 281 .
LU G F , JIN Z , ZOU J . Face recognition using discriminant sparsity neighborhood preserving embedding [J]. Knowledge-Based Systems , 2012 , 31 : 119 - 127 .
XU D , YAN S C , TAO D C , et al . Marginal fisher analysis and its variants for human gait recognition and content- based image retrieval [J]. IEEE Transactions on Image Processing , 2007 , 16 ( 11 ): 2811 - 2821 .
LUO F L , HUANG H , DUAN Y L , et al . Local geometric structure feature for dimensionality reduction of hyperspectral imagery [J]. Remote Sensing , 2017 , 9 ( 8 ): 790 .
DAI X A , CHENG J Y , GAO Y , et al . Deep belief network for feature extraction of urban artificial targets [J]. Mathematical Problems in Engineering , 2020 , 2020 : 1 - 13 .
LI Z Y , HUANG H , DUAN Y L , et al . DLPNet: a deep manifold network for feature extraction of hyperspectral imagery [J]. Neural Networks , 2020 , 129 : 7 - 18 .
ZHANG H K , LI Y , JIANG Y N . Deep learning for hyperspectral imagery classification: the state of the art and prospects [J]. Acta Automatica Sinica , 2018 , 44 ( 6 ): 961 - 977 . (in Chinese)
RATLE F , CAMPS-VALLS G , WESTON J . Semisupervised neural networks for efficient hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2010 , 48 ( 5 ): 2271 - 2282 .
CHEN Y S , LIN Z H , ZHAO X , et al . Deep learning-based classification of hyperspectral data [J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 2014 , 7 ( 6 ): 2094 - 2107 .
LI Z Y , HUANG H , LI Y , et al . M3DNet: a manifold-based discriminant feature learning network for hyperspectral imagery [J]. Expert Systems With Applications , 2020 , 144 : 113089 .
ZHOU X C , LIU N , TANG F , et al . A deep manifold learning approach for spatial-spectral classification with limited labeled training samples [J]. Neurocomputing , 2019 , 331 : 138 - 149 .
ZHANG X B , WANG X M , WANG K , et al . Recognition of the Martian minerals based on the deep belief networks method: Application in the CRISM images [J]. Bulletin of Geological Science and Technology , 2020 , 39 ( 4 ): 189 - 200 . (in Chinese)
GADER P , ZARE A , CLOSE R , et al . “ MUUFL Gulfport Hyperspectral and LiDAR Airborne Data Set ,” University of Florida , Gainesville, FL, Tech. Rep. REP- 2013 -570, Oct. 2013.
DIAKITE A , GUI J S , FU X P . Hyperspectral image classification using 3D 2D CNN [J]. IET Image Processing , 2021 , 15 ( 5 ): 1083 - 1092 .