Hyperspectral Multilevel GCN and CNN Feature Fusion for Change Detection

Abstract

Hyperspectral image (HSI) change detection focuses on identifying differences in multitemporal HSIs. Graph convolutional networks (GCNs) have demonstrated greater promise than convolutional neural networks (CNNs) in remote sensing, particularly for processing HSIs. This is due to GCN's ability to handle non-Euclidean graph-structured information, as opposed to the fixed kernel operations of CNN based on Euclidean structures. Specifically, GCN operates predominantly on superpixel-based nodes. This article proposes a method, named hyperspectral multilevel GCN and CNN feature fusion (HMGCF) for change detection, that integrates superpixel-level GCN with pixel-level CNN for feature extraction and efficient change detection in HSI. The proposed method utilizes the strengths of both CNN and GCN; the CNN branch focuses on feature learning in small-scale, regular regions, while the GCN branch handles large-scale, irregular regions. This approach generates complementary spectral-spatial features at both pixel and superpixel levels. To bridge the structural incompatibility between the Euclidean-data-oriented CNN and the non-Euclidean-data-oriented GCN, HMGCF introduces a graph encoder and decoder. These elements help in propagating features between image pixels and graph nodes, allowing CNN and GCN to function within an integrated end-to-end framework. HMGCF integrates graph encoding into the network, edge weights, and node representations from training data. Ablation studies on four datasets reveal that the combination of CNN and GCN branches in the HMGCF model consistently outperforms existing methods by margins ranging from 0.5% to 2.5%. In addition, HMGCF shows significant improvements in both kappa and F1 scores in all datasets.