Convolutional neural networks for hyperspectral image classification

By Shiqi Yu, Sen Jia, Chunyan Xu. Neurocomputing, 2017

609 citations recorded by the source corpus.

As a powerful visual model, convolutional neural networks (CNNs) have demonstrated remarkable performance in various visual recognition problems, and attracted considerable attention in recent years. However, due to the highly correlated bands and insufficient training samples of hyperspectral image data, it still remains a challenging problem to effectively apply the CNN models on hyperspectral images. In this paper, an efficient CNN architecture has been proposed to boost its discriminative capability for hyperspectral image classification, in which the original data is used as the input and the final CNN outputs are the predicted class-related results. The proposed CNN infrastructure has several distinct advantages. Firstly, different from traditional classification methods those need hand-crafted features, the CNN model used here is designed to deal with the problem of hyperspectral image analysis in an end-to-end way. Secondly, the parameters of the CNN model are optimized from a small training set, while the over-fitting problem of the neural network has been alleviated to some extent. Finally, in order to better deal with the hyperspectral image information, 1 ? 1 convolutional layers have been adopted, and an average pooling layer and larger dropout rates have also been employed in the whole CNN procedure. The experiments on three benchmark data sets have demonstrated that the proposed CNN architecture considerably outperforms other state-of-the-art methods.

Subfields: Atmospheric Science, Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/ec9aa46ebc50a03ea9d7d20d80a232e6bd4293de

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Atmospheric Science

1. Pedram Ghamisi
2. Danfeng Hong
3. Qian Du
4. Wei Li
5. Lichuan Gu
6. Xiaohui Yuan
7. Jianfang Shi
8. Sen Jia

In Computer Vision and Pattern Recognition

1. Danfeng Hong
2. Jocelyn Chanussot
3. Swalpa Kumar Roy
4. Jing Yao
5. Qian Du
6. Xin Wu
7. Bing Zhang
8. Pedram Ghamisi

In Media Technology

1. Danfeng Hong
2. Jocelyn Chanussot
3. Qian Du
4. Swalpa Kumar Roy
5. Jing Yao
6. Pedram Ghamisi
7. Bing Zhang
8. Xin Wu

Citations in the corpus

• Classification of Hyperspectral and LiDAR Data Using Coupled CNNs (2020, background): “It has been proved that CNNs are capable of extracting high-level features, which are usually invariant to the nonlinearities of hyperspectral [34]–[36] and LiDAR data [30], [37].”
• A review of deep learning methods for semantic segmentation of remote sensing imagery (2021, methodology): “…as the loss function (Alshehhi et al., 2017; Cheng et al., 2017; Fu et al., 2017; Kussul et al., 2017; Li, Zhang, & Shen, 2017; Wang et al., 2017; Yu, Jia, & Xu, 2017; Chen et al., 2018; Chen et al., 2018; Chen et al., 2018; Kestur et al., 2018; Liu et al., 2018; Liu, Yu, Yu, & Wan, 2018; Pan et…”
• Multisource Remote Sensing Data Classification Based on Convolutional Neural Network (2018, methodology): “In [41]–[43], more CNNs were employed for deep feature extraction.”
• Recent Advances in Convolutional Neural Network Acceleration (2018, background): “Therefore, it shows excellent performance in solving computer vision problems such as image classification, recognition and understanding [1, 2, 3].”
• A survey: Deep learning for hyperspectral image classification with few labeled samples (2021, methodology): “Classification accuracy obtained by S-DMM [121], 3DCAE [40], SSDL [37], TwoCnn [122], 3DVSCNN [123], SSLstm [75], CNN HSI [28] and SAE LR [34] on PaviaU.”
• Hyperspectral Image Classification—Traditional to Deep Models: A Survey for Future Prospects (2021, result): “Similar to [137], [131] also utilized the global average pooling”