Deep Convolutional Neural Networks for Hyperspectral Image Classification

By Wei Hu, Yangyu Huang, Wei Li, Fan Zhang, Hengchao Li. J. Sensors, 2015

1,704 citations recorded by the source corpus.

Recently, convolutional neural networks have demonstrated excellent performance on various visual tasks, including the classification of common two-dimensional images. In this paper, deep convolutional neural networks are employed to classify hyperspectral images directly in spectral domain. More specifically, the architecture of the proposed classifier contains five layers with weights which are the input layer, the convolutional layer, the max pooling layer, the full connection layer, and the output layer. These five layers are implemented on each spectral signature to discriminate against others. Experimental results based on several hyperspectral image data sets demonstrate that the proposed method can achieve better classification performance than some traditional methods, such as support vector machines and the conventional deep learning-based methods.

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

View Paper: https://www.semanticscholar.org/paper/2369db9921078c4bb76072ef7d6426e9f1dbfdb5

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

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

MambaHSI: Spatial–Spectral Mamba for Hyperspectral Image Classification (2025, background): “Hu et al. [43] extracts spectral features via a five-layer 1-”
Deep Learning for Remote Sensing Data: A Technical Tutorial on the State of the Art (2016, background): “In the RS community, there are also some recent works on CNN-based RS image pixel classification [45]–[47], target recognition [48], [49], and scene understanding [50].”
Multi-scale 3D deep convolutional neural network for hyperspectral image classification (2017, methodology): “Hu et al.[7] proposed a two layer 1D CNN architecture to extract spectral features and achieve significant results.”
Feature Extraction for Hyperspectral Imagery: The Evolution From Shallow to Deep: Overview and Toolbox (2020, methodology): “For hyperspectral images, CNNs can be used for extracting spectral features [99] or spatial features [100]–[102], depending on the inputs of networks.”
3-D Deep Learning Approach for Remote Sensing Image Classification (2018, methodology): “This approach is close to the one presented in [41] but adds spatial information.”
Multisource Remote Sensing Data Classification Based on Convolutional Neural Network (2018, methodology): “To illustrate the performance of the proposed two-channel CNN for multisource remote sensing data classification, we compared with several traditional classifiers, such as SVM, ELM, and the recently developed CNN pixel-pair features (CNN-PPF) [31], [38], [39].”
Deep Learning for Hyperspectral Image Classification: An Overview (2019, methodology): “In addition, 1-D CNN [55]–[58], 1-D GAN [46], [59], and RNN [44], [58], [60] were also used to extract spectral features for HSI classification.”
Spectral-Spatial Attention Networks for Hyperspectral Image Classification (2019, neutral): “Deep learning algorithms have been introduced to modern HSI analysis due to their outstanding predictive power, and they can extract more discriminative features and achieve a better performance than traditional shallow classifiers [2,12].”
Diverse Region-Based CNN for Hyperspectral Image Classification (2018, methodology): “Computational complexity of training and testing procedures using the proposed DR-CNN, CNN [33], and CNN-PPF [38] is summarized in Table XV.”
Hyperspectral Image Classification—Traditional to Deep Models: A Survey for Future Prospects (2021, methodology): “21: Classification Maps obtained by MLR [308], SVM [309], MLP [310], RNN [59], LSTM [311], GRU [312], CNN-1D [313],”
Spatial-Spectral Transformer for Hyperspectral Image Classification (2021, methodology): “proposed 1-D CNN with five convolutional layers to extract the spectral features of HSI [29].”
Learning to Diversify Deep Belief Networks for Hyperspectral Image Classification (2017, methodology): “Eight classes have only few labeled training samples and thus were discarded to make the experimental analysis more statistically significant [21].”