基于超像素分割的图注意力网络的高光谱图像分类

吉林大学学报(理学版) ›› 2024, Vol. 62 ›› Issue (2): 357-0368.

基于超像素分割的图注意力网络的高光谱图像分类

高路尧^1,2, 胡长虹¹, 肖树林^1,2

1. 中国科学院长春光学精密机械与物理研究所, 长春 130033； 2. 中国科学院大学, 北京 100049

收稿日期:2023-05-04 出版日期:2024-03-26 发布日期:2024-03-26
通讯作者: 胡长虹 E-mail:changhonghu@rocketmail.com

Hyperspectral Image Classification Based on Superpixel Segmentation with Graph Attention Networks

GAO Luyao^1,2, HU Changhong¹, XIAO Shulin^1,2

1. Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, China；
2. University of Chinese Academy of Sciences, Beijing 100049, China

Received:2023-05-04 Online:2024-03-26 Published:2024-03-26

摘要/Abstract

摘要： 针对卷积神经网络(CNN）仅能应用于欧氏数据, 无法有效获取像素间的全局关系特征以及长距离上下文信息的问题, 构建一个基于超像素分割的图注意力网络SSGAT. 该网络将超像素分割后的超像素块视为图结构中的图节点, 有效减少了图结构的复杂度, 并降低了分类图的噪声. 在3个数据集上对SSGAT及对比算法的分类精度进行测试, 分别获得了94.11%,95.22%,96.37%的总体分类精度. 结果表明该方法性能优异, 在处理大尺度区域的分类问题时优势明显.

关键词: 高光谱图像, 图注意力网络, 残差机制, 超像素分割

Abstract: Aiming at the problem that convolutional neural network (CNN) could only be applied to Euclidean data and could not effectively
obtain global relationship features between pixels and long-distance contextual information, we constructed a superpixel segmentation-based graph attention network (SSGAT). The network treated the segmented superpixel blocks as graph nodes in the graph structure, effectively reducing the complexity of the graph structure and reducing the noise of the classification graph.
The classification accuracy of SSGAT and the comparison algorithm were tested on three datasets, and overall classification accuracy of 94.11%, 95.22%, and 96.37% were obtained, respectively. The results show that the method has excellent performance and significant advantages in dealing with classification problems in large-scale regions.

Key words: hyperspectral image, graph attention network, residual mechanism, superpixel segmentation

中图分类号:

TP391

高路尧, 胡长虹, 肖树林. 基于超像素分割的图注意力网络的高光谱图像分类[J]. 吉林大学学报(理学版), 2024, 62(2): 357-0368.

GAO Luyao, HU Changhong, XIAO Shulin. Hyperspectral Image Classification Based on Superpixel Segmentation with Graph Attention Networks[J]. Journal of Jilin University Science Edition, 2024, 62(2): 357-0368.

[1]	周丰丰, 张金楷. 具有局部和全局注意力机制的图注意力网络学习单样本组学数据表征[J]. 吉林大学学报(理学版), 2023, 61(6): 1351-1357.
[2]	周航, 苏延池, 李占山, 花昀峤. 基于子空间表示和加权低秩张量正则化的高光谱图像混合噪声去除方法[J]. 吉林大学学报(理学版), 2023, 61(1): 118-126.
[3]	蔡旭航, 朱留存, 张震, 张恒艳, 郑晓东. 基于多尺度超像素融合的RGB-D单幅图像阴影检测算法[J]. 吉林大学学报(理学版), 2022, 60(2): 392-400.
[4]	任伟建, 刘泽宇, 霍凤财, 康朝海, 任璐, 张永丰. 一种改进的多光谱遥感图像超像素分割算法[J]. 吉林大学学报(理学版), 2022, 60(2): 351-360.