A Mutually Enhancement Network for Superpixel Segmentation and Classification of Hyperspectral Image
- 11 September 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 63 (01962892) , 1-14
Abstract
Most existing hyperspectral image (HSI) classification methods primarily focus on capturing subtle spectral variations by leveraging local spectral–spatial cues derived from patch-level representations. However, limited attention has been given to exploring the global spatial contextual correlations among pixels of HSI. In this study, we propose the superpixel segmentation and classification mutual enhancement network ( $\mathbf {S}^{2}$ CMEN), a novel framework that integrates global spatial correlations with spectral information through the mutual enhancement of superpixel segmentation and classification. Specifically, a global spatial adaptive module (GSAM) is designed to obtain the direct correlation of the global classes in HSI. It consists of an adaptive spectral-superpixel network (ASSN) and a graph convolutional network (GCN), forming a synergistic architecture that effectively captures global spatial relationships by adaptively deriving superpixel results from HSIs. Notably, GSAM offers a transferable global spatial representation for HSI tasks, enabling integration with other spectral feature extraction models. Furthermore, we develop a spatial–spectral fusion module (SSFM) to obtain comprehensive spectral features and fuse them with the extracted global spatial features. Finally, under the constraint of a unit loss, the mutual enhancement strategy (MES) can make the superpixel segmentation loss and the classification loss mutually enhance each other for better performance. We conducted extensive experiments on three public datasets. The proposed $\mathbf {S}^{2}$ CMEN achieves overall classification accuracies of 97.38%, 92.33%, and 91.38% on Indian Pines (IPs), Pavia University (PU), and Houston, respectively, consistently surpassing existing state-of-the-art methods.Keywords
Funding Information
- National Natural Science Foundation of China (62076143, 62173193)
- Fundamental research promotion plan of Qilu University of Technology (2021JC02009)
- Natural Science Foundation of Shandong province, China (ZR2022MF237)
This publication has 43 references indexed in Scilit:
- SuperFusion: A Versatile Image Registration and Fusion Network with Semantic AwarenessIEEE/CAA Journal of Automatica Sinica, 2022
- SwinFusion: Cross-domain Long-range Learning for General Image Fusion via Swin TransformerIEEE/CAA Journal of Automatica Sinica, 2022
- Swin Transformer: Hierarchical Vision Transformer using Shifted Windows2021
- A comprehensive and fast microplastics identification based on near-infrared hyperspectral imaging (HSI-NIR) and chemometricsEnvironmental Pollution, 2021
- Superpixel Segmentation With Fully Convolutional Networks2020
- Hyperspectral imaging in crop fields: precision agriculture2019
- Superpixel Sampling Networks2018
- Hyperspectral image denoising with superpixel segmentation and low-rank representationInformation Sciences, 2017
- Deep Learning-Based Classification of Hyperspectral DataIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014
- Target detection based on a dynamic subspacePattern Recognition, 2014