Deep hierarchical spectral-spatial feature fusion for hyperspectral image classification based on convolutional neural network

被引:0
作者
Bera, Somenath [1 ]
Varish, Naushad [2 ]
Yaqoob, Syed irfan [3 ]
Rafi, Mudassir [4 ,5 ]
Shrivastava, Vimal K. [6 ]
机构
[1] Nirma Univ, Inst Technol, Sch Technol, Dept Comp Sci & Engn, Ahmadabad, India
[2] GITAM Univ, Comp Sci & Engn, Telangana, India
[3] Chandigarh Univ, Comp Sci & Engn AIT, Ajitgarh, Punjab, India
[4] SRM Univ, Comp Sci & Engn, Amaravati, Andhra Prades, India
[5] King Khalid Univ, Coll Comp Sci, Dept Comp Sci, Abha, Saudi Arabia
[6] Kalinga Inst Ind Technol KIIT, Sch Elect Engn, Bhubaneswar, India
关键词
CNN; deep learning; feature fusion; feature extraction; hyperspectral image classification; informative bands; AUGMENTATION; ARCHITECTURE; EXTRACTION;
D O I
10.3233/IDA-230927
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Joint spectral-spatial feature extraction has been proven to be the most effective part of hyperspectral image (HSI) classification. But, due to the mixing of informative and noisy bands in HSI, joint spectral-spatial feature extraction using convolutional neural network (CNN) may lead to information loss and high computational cost. More specifically, joint spectral-spatial feature extraction from excessive bands may cause loss of spectral information due to the involvement of convolution operation on non-informative spectral bands. Therefore, we propose a simple yet effective deep learning model, named deep hierarchical spectral-spatial feature fusion (DHSSFF), where spectral-spatial features are exploited separately to reduce the information loss and fuse the deep features to learn the semantic information. It makes use of abundant spectral bands and few informative bands of HSI for spectral and spatial feature extraction, respectively. The spectral and spatial features are extracted through 1D CNN and 3D CNN, respectively. To validate the effectiveness of our model, the experiments have been performed on five well-known HSI datasets. Experimental results demonstrate that the proposed method outperforms other state-of-the-art methods and achieved 99.17%, 98.84%, 98.70%, 99.18%, and 99.24% overall accuracy on Kennedy Space Center, Botswana, Indian Pines, University of Pavia, and Salinas datasets, respectively.
引用
收藏
页码:385 / 407
页数:23
相关论文
共 47 条
[41]   Hyperspectral Image Classification Method Based on CNN Architecture Embedding With Hashing Semantic Feature [J].
Yu, Chunyan ;
Zhao, Meng ;
Song, Meiping ;
Wang, Yulei ;
Li, Fang ;
Han, Rui ;
Chang, Chein-I .
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2019, 12 (06) :1866-1881
[42]   An eye detection method based on convolutional neural networks and support vector machines [J].
Yu, Mingxin ;
Tang, Xiaoying ;
Lin, Yingzi ;
Schmidt, David ;
Wang, Xiangzhou ;
Guo, Yikang ;
Liang, Bo .
INTELLIGENT DATA ANALYSIS, 2018, 22 (02) :345-362
[43]   Deep learning in remote sensing scene classification: a data augmentation enhanced convolutional neural network framework [J].
Yu, Xingrui ;
Wu, Xiaomin ;
Luo, Chunbo ;
Ren, Peng .
GISCIENCE & REMOTE SENSING, 2017, 54 (05) :741-758
[44]   Adaptive total variation-based spectral-spatial feature extraction of hyperspectral image [J].
Zhang, Guoyun ;
Wang, Jinping ;
Zhang, Xiaofei ;
Fei, Hongyan ;
Tu, Bing .
JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION, 2018, 56 :150-159
[45]   Spectral-spatial classification of hyperspectral imagery using a dual-channel convolutional neural network [J].
Zhang, Haokui ;
Li, Ying ;
Zhang, Yuzhu ;
Shen, Qiang .
REMOTE SENSING LETTERS, 2017, 8 (05) :438-447
[46]   Spectral-Spatial Feature Extraction for Hyperspectral Image Classification: A Dimension Reduction and Deep Learning Approach [J].
Zhao, Wenzhi ;
Du, Shihong .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2016, 54 (08) :4544-4554
[47]   Spectral-Spatial Exploration for Hyperspectral Image Classification via the Fusion of Fully Convolutional Networks [J].
Zou, Liang ;
Zhu, Xingliang ;
Wu, Changfeng ;
Liu, Yong ;
Qu, Lei .
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2020, 13 (13) :659-674