An Efficient Topology Construction Scheme Designed for Graph Neural Networks in Hyperspectral Image Classification

Superpixel-based Graph Neural Networks (GNNs) have achieved remarkable success in hyperspectral image (HSI) classification tasks, primarily due to their ability to capture the implicit topological structure in the data while maintaining low computational complexity by propagating information between...

Full description

Saved in:
Bibliographic Details
Main Authors: Yu Zhang, Xin Li, Yaoqun Xu
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11113241/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849340445776150528
author Yu Zhang
Xin Li
Yaoqun Xu
author_facet Yu Zhang
Xin Li
Yaoqun Xu
author_sort Yu Zhang
collection DOAJ
description Superpixel-based Graph Neural Networks (GNNs) have achieved remarkable success in hyperspectral image (HSI) classification tasks, primarily due to their ability to capture the implicit topological structure in the data while maintaining low computational complexity by propagating information between spatially adjacent superpixels. However, the assumption of treating pixel features within superpixels as identical representations may limit the model&#x2019;s expressive power. This is because land cover regions in HSI are often irregular, which leads to a difficult-to-resolve contradiction between the superpixel segmentation scale and the homogeneity of pixel labels within superpixels. To fundamentally address this issue, we reinterpreted the implicit topological structure between pixels on the basis of their spectral feature similarity and spatial position dependencies. Considering computational bottlenecks, we proposed a new subgraph partitioning method and sorting selection technique to choose important relationships from the graph, thereby constructing an a priori topology conducive to downstream tasks on HSI of any scale. Based on this a priori topology, we designed a GCN model for learning pixel feature representations and integrated it into a unified framework. We conducted comprehensive experiments on three widely used benchmark datasets. The results show that, compared to the mainstream superpixel-level GCN models, the proposed method improved the overall accuracy (OA), the average accuracy (AA), and the Kappa coefficient (Kappa) by an average of 2.0%, 5.4%, and 2.3%, respectively, across the three datasets. Moreover, on some datasets, our method outperformed several recent multi-scale feature fusion models. We also observed that different models exhibit different performances when dealing with land cover areas with different characteristics. In particular, by combining these models with our method, the classification performance was significantly improved. Our code is open-source on the public platform GitHub: <uri>https://github.com/LittleBlackBearLiXin/An-Efficient-Topology-Construction-Scheme-Designed-for-Graph-Neural-Networks</uri>
format Article
id doaj-art-0e98f08d72e64219a090d8daf0cf654d
institution Kabale University
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-0e98f08d72e64219a090d8daf0cf654d2025-08-20T03:43:55ZengIEEEIEEE Access2169-35362025-01-011314083014084610.1109/ACCESS.2025.359599711113241An Efficient Topology Construction Scheme Designed for Graph Neural Networks in Hyperspectral Image ClassificationYu Zhang0https://orcid.org/0009-0007-7276-9029Xin Li1https://orcid.org/0009-0007-8527-9467Yaoqun Xu2https://orcid.org/0000-0002-5047-9350School of Computer Science and Information Engineering, Harbin University of Commerce, Harbin, Heilongjiang, ChinaSchool of Computer Science and Information Engineering, Harbin University of Commerce, Harbin, Heilongjiang, ChinaSchool of Computer Science and Information Engineering, Harbin University of Commerce, Harbin, Heilongjiang, ChinaSuperpixel-based Graph Neural Networks (GNNs) have achieved remarkable success in hyperspectral image (HSI) classification tasks, primarily due to their ability to capture the implicit topological structure in the data while maintaining low computational complexity by propagating information between spatially adjacent superpixels. However, the assumption of treating pixel features within superpixels as identical representations may limit the model&#x2019;s expressive power. This is because land cover regions in HSI are often irregular, which leads to a difficult-to-resolve contradiction between the superpixel segmentation scale and the homogeneity of pixel labels within superpixels. To fundamentally address this issue, we reinterpreted the implicit topological structure between pixels on the basis of their spectral feature similarity and spatial position dependencies. Considering computational bottlenecks, we proposed a new subgraph partitioning method and sorting selection technique to choose important relationships from the graph, thereby constructing an a priori topology conducive to downstream tasks on HSI of any scale. Based on this a priori topology, we designed a GCN model for learning pixel feature representations and integrated it into a unified framework. We conducted comprehensive experiments on three widely used benchmark datasets. The results show that, compared to the mainstream superpixel-level GCN models, the proposed method improved the overall accuracy (OA), the average accuracy (AA), and the Kappa coefficient (Kappa) by an average of 2.0%, 5.4%, and 2.3%, respectively, across the three datasets. Moreover, on some datasets, our method outperformed several recent multi-scale feature fusion models. We also observed that different models exhibit different performances when dealing with land cover areas with different characteristics. In particular, by combining these models with our method, the classification performance was significantly improved. Our code is open-source on the public platform GitHub: <uri>https://github.com/LittleBlackBearLiXin/An-Efficient-Topology-Construction-Scheme-Designed-for-Graph-Neural-Networks</uri>https://ieeexplore.ieee.org/document/11113241/Hyperspectral image classificationgraph neural networksmachine learning
spellingShingle Yu Zhang
Xin Li
Yaoqun Xu
An Efficient Topology Construction Scheme Designed for Graph Neural Networks in Hyperspectral Image Classification
IEEE Access
Hyperspectral image classification
graph neural networks
machine learning
title An Efficient Topology Construction Scheme Designed for Graph Neural Networks in Hyperspectral Image Classification
title_full An Efficient Topology Construction Scheme Designed for Graph Neural Networks in Hyperspectral Image Classification
title_fullStr An Efficient Topology Construction Scheme Designed for Graph Neural Networks in Hyperspectral Image Classification
title_full_unstemmed An Efficient Topology Construction Scheme Designed for Graph Neural Networks in Hyperspectral Image Classification
title_short An Efficient Topology Construction Scheme Designed for Graph Neural Networks in Hyperspectral Image Classification
title_sort efficient topology construction scheme designed for graph neural networks in hyperspectral image classification
topic Hyperspectral image classification
graph neural networks
machine learning
url https://ieeexplore.ieee.org/document/11113241/
work_keys_str_mv AT yuzhang anefficienttopologyconstructionschemedesignedforgraphneuralnetworksinhyperspectralimageclassification
AT xinli anefficienttopologyconstructionschemedesignedforgraphneuralnetworksinhyperspectralimageclassification
AT yaoqunxu anefficienttopologyconstructionschemedesignedforgraphneuralnetworksinhyperspectralimageclassification
AT yuzhang efficienttopologyconstructionschemedesignedforgraphneuralnetworksinhyperspectralimageclassification
AT xinli efficienttopologyconstructionschemedesignedforgraphneuralnetworksinhyperspectralimageclassification
AT yaoqunxu efficienttopologyconstructionschemedesignedforgraphneuralnetworksinhyperspectralimageclassification