Abstract:Convolutional neural networks (CNNs) and graph convolutional networks (GCNs) have led to promising advancements in hyperspectral image (HSI) classification; however, traditional CNNs with fixed square convolution kernels are insufficiently flexible to handle irregular structures. Similarly, GCNs that employ superpixel nodes instead of pixel nodes may overlook pixel-level features; both networks tend to extract features locally and cause loss of multilayer contextual semantic information during feature extraction due to the fixed kernel. To leverage the strengths of CNNs and GCNs, we propose a multiscale pixel-level and superpixel-level (MPAS)-based HSI classification method. The network consists of two sub-networks for extracting multi-level information of HSIs: a multi-scale hybrid spectral–spatial attention convolution branch (HSSAC) and a parallel multi-hop graph convolution branch (MGCN). HSSAC comprehensively captures pixel-level features with different kernel sizes through parallel multi-scale convolution and cross-path fusion to reduce the semantic information loss caused by fixed convolution kernels during feature extraction and learns adjustable weights from the adaptive spectral–spatial attention module (SSAM) to capture pixel-level feature correlations with less computation. MGCN can systematically aggregate multi-hop contextual information to better model HSIs' spatial background structure using the relationship between parallel multi-hop graph transformation nodes. The proposed MPAS effectively captures multi-layer contextual semantic features by leveraging pixel-level and superpixel-level spectral–spatial information, which improves the performance of the HSI classification task while ensuring computational efficiency. Extensive evaluation experiments on three real-world HSI datasets demonstrate that MPAS outperforms other state-of-the-art networks, demonstrating its superior feature learning capabilities.

Multiscale Spectral-Spatial Hyperspectral Image Classification with Adaptive Filtering

Attention in Attention for Hyperspectral with High Spatial Resolution (H) Image Classification

Spectral-Spatial Classification of Hyperspectral Image Using Improved Functional Principal Component Analysis

Hyperspectral Classification Based on Coupling Multiscale Super-Pixels and Spatial Spectral Features

Subpixel Multilevel Scale Feature Learning and Adaptive Attention Constraint Fusion for Hyperspectral Image Classification

Effects of calcium channel blockers on the contractile response to dihydroergotamine in isolated human femoral veins

Hyperspectral Image Classification based on Multi-Scale Convolutional Features and Multi-Attention Mechanisms

Subpixel Mapping of Hyperspectral Image Based on Multiscale and Multifeature

A Feature Embedding Network with Multiscale Attention for Hyperspectral Image Classification

Superpixel-Based Multiple Statistical Feature Extraction Method for Classification of Hyperspectral Images

Spatial–Spectral Adaptive Learning With Pixelwise Filtering for Hyperspectral Image Classification

Multiscale Densely Connected Attention Network for Hyperspectral Image Classification

MBSFC: hyperspectral image classification based on multi-branch and spectral feature conversion

A Spectral-Spatial Classification of Hyperspectral Images Based on the Algebraic Multigrid Method and Hierarchical Segmentation Algorithm

Classification of Hyperspectral Imagery Based on Spectral Gradient, SVM and Spatial Random Forest

Hyperspectral image classification based on multiscale piecewise spectral-spatial attention network

MSFF: A Multi-Scale Feature Fusion Convolutional Neural Network for Hyperspectral Image Classification

Multi-scale guided feature extraction and classification algorithm for hyperspectral images

Multiscale Pixel-Level and Superpixel-Level Method for Hyperspectral Image Classification: Adaptive Attention and Parallel Multi-Hop Graph Convolution

Adaptive Spatial-Spectral Feature Learning for Hyperspectral Image Classification.

Multiscale Residual Weakly Dense Network with Attention Mechanism for Hyperspectral Image Classification