S<sup>2</sup>PNet: An Interactive Learning Framework for Addressing Spatial&#x2013;Spectral Heterogeneity in H<sup>2</sup> Imagery Classification

Hyperspectral imagery with high spatial resolution (H<sup>2</sup>) imagery can synchronously obtain the spectral and spatial features of objects, thus providing richer information. However, the exacerbated spatial&#x2013;spectral heterogeneity poses new challenges for classification....

Full description

Saved in:
Bibliographic Details
Main Authors: Shuai Zhang, Yonghua Jiang, Chengjun Wang, Meilin Tan, Bin Du, Feng Tian
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10684570/
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Hyperspectral imagery with high spatial resolution (H<sup>2</sup>) imagery can synchronously obtain the spectral and spatial features of objects, thus providing richer information. However, the exacerbated spatial&#x2013;spectral heterogeneity poses new challenges for classification. In this study, an interactive learning framework was proposed to address the current issues in H<sup>2</sup> imagery classification. Specifically, we propose a spectral&#x2013;spatial purification network (S<sup>2</sup>PNet) to improve classification accuracy. First, a multistage spectral purification module is designed to purify noisy information and mitigate spectral heterogeneity, achieving interaction between spectral optimization and classification. Second, a global&#x2013;local mutual guide module is utilized to realize image&#x2013;pixel-level feature interaction, thus enhancing the spatial discriminability of extracted features and reducing spatial heterogeneity. Third, the introduction of dual-stream semantic progressive module facilitates shallow-deep feature interaction, reducing the semantic gap in internal network and enabling a smoother information flow. We validated our approach using the public WHU-Hi hyperspectral datasets and large-scale Houston datasets. Experimental results demonstrate that S<sup>2</sup>PNet achieves the highest classification accuracy across all tests, significantly outperforming state-of-the-art methods.
ISSN:1939-1404
2151-1535