Spectral–Spatial Feature Extraction for Hyperspectral Image Classification: A Dimension Reduction and Deep Learning Approach

By Wenzhi Zhao, Shihong Du. IEEE Transactions on Geoscience and Remote Sensing, 2016

1,136 citations recorded by the source corpus.

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

View Paper: https://www.semanticscholar.org/paper/4cf28fc05744d9cfd3206b6b6e8a81c823e740f6

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

Learning and Transferring Deep Joint Spectral–Spatial Features for Hyperspectral Classification (2017, methodology): “In [26], CNN was used to extract spatial features from HSI, and manifold learning was used to extract spectral features.”
Deep learning in remote sensing: a review (2017, methodology): “To avoid overfitting, Zhao and Du [28] suggest a spectral-spatial-feature-based classification framework, which jointly makes use of a local-discriminant embedding-based dimension-reduction algorithm and a 2-D CNN.”
Learning to Diversify Deep Belief Networks for Hyperspectral Image Classification (2017, methodology): “In addition, Zhao and Du [23] also aimed to use the spatial and spectral information but with a separated way, where a local discriminant embedding algorithm was proposed for spectral feature extraction and then a CNN was used to extract spatial-related features at high levels.”
Deep learning for remote sensing image classification: A survey (2018, methodology): “In Zhao and Du (2016), based on local discriminant embedding (LDE) (Chen, Chang, & Liu, 2005), a balanced LDE method was proposed and jointly used with a 2D-CNN to obtain the final classification result.”
A review of deep learning methods for semantic segmentation of remote sensing imagery (2021, methodology): “Zhao and Du (2016) proposed a multiscale convolutional neural network (MCNN) to learn deep features of spatial relationships.”
Hyperspectral Imaging: A Review on UAV-Based Sensors, Data Processing and Applications for Agriculture and Forestry (2017, background): “For example, in [118], a framework that joins dimension reduction and deep learning to allow hyperspectral image”
Feature Extraction for Hyperspectral Imagery: The Evolution From Shallow to Deep: Overview and Toolbox (2020, methodology): “In [100], [101], [103], PCA was employed to extract the first principal components of hyperspectral images as inputs of CNNs, thus simplifying the network structures.”
Multiscale Dynamic Graph Convolutional Network for Hyperspectral Image Classification (2019, background): “Recently, the convolutional neural network (CNN) has emerged as a powerful tool for hyperspectral image classification [25]–[27].”
Spectral–Spatial Feature Tokenization Transformer for Hyperspectral Image Classification (2022, methodology): “To demonstrate the effectiveness of the proposed model, several representative methods were selected for comparative experiments: SVM, EMAP [30], 1-D-CNN [36], 2-D-CNN [37], 3-D-CNN [40], SSRN [42], Cubic-CNN [45], HybridSN [41], and the proposed SSFTT method.”
Spectral-Spatial Classification of Hyperspectral Imagery with 3D Convolutional Neural Network (2017, neutral): “To address this important issue, a more sophisticated procedure for additional spectral feature extraction can be employed as reported in [32].”
Recent Advances on Spectral–Spatial Hyperspectral Image Classification: An Overview and New Guidelines (2018, methodology): “References [68] and [69] used CNN strategy to fulfill the classification where feature layer and classification layer were trained jointly and the training was usually not”
Advanced Multi-Sensor Optical Remote Sensing for Urban Land Use and Land Cover Classification: Outcome of the 2018 IEEE GRSS Data Fusion Contest (2019, background): “Following the great success of deep learning in computer vision ﬁeld [19]–[21], many deep models have been proposed to address the remote sensing image classiﬁcation task [22]–[28].”
Spectral–Spatial Residual Network for Hyperspectral Image Classification: A 3-D Deep Learning Framework (2018, methodology): “For example, [23] used CNNs to extract spatial features, which were integrated with spectral features that learned from balanced local discriminant embedding, for HSI classification.”
Explainable artificial intelligence: a comprehensive review (2021, background): “For example, some studies proposed compressing the deep learning models 1 3 from a high-dimensional representation into a low-dimensional representation (Zhao and Du 2016; Erfani et al.”
Going Deeper With Contextual CNN for Hyperspectral Image Classification (2016, methodology): “Accordingly, the reduction of the spectral dimension of the hyperspectral images is in general initially performed to ﬁt the input data into the small-scale networks by using techniques, such as principal component analysis (PCA) [9], balanced local discriminant embedding (BLDE) [3], pairwise constr”
Deep Learning for Classification of Hyperspectral Data: A Comparative Review (2019, positive): “Trying to go around this problem, [51] and [52] proposed another approach.”