帮助 关于我们

返回检索结果

基于层级Transformer的高光谱图像分类方法
Hyperspectral image classification method based on hierarchical transformer network

查看参考文献30篇

张艺超 1,2   郑向涛 1,3 *   卢孝强 1,3  
文摘 高光谱图像分类将每个像素分类至预设的地物类别,对环境测绘等各类地球科学任务有着至关重要的意义。近年来,学者们尝试利用深度学习框架进行高光谱图像分类,取得了令人满意的效果。然而这些方法在光谱特征的提取上仍存在一定缺陷。本文提出一个基于自注意力机制的层级融合高光谱图像分类框架(hierarchical self-attention network,HSAN)。首先,构建跳层自注意力模块进行特征学习,利用Transformer结构中的自注意力机制捕获上下文信息,增强有效信息贡献。然后,设计层级融合方式,进一步缓解特征学习过程中的有效信息损失,增强各层级特征联动。在Pavia University及Houston2013数据集上的试验表明,本文提出的框架相较于其他高光谱图像分类框架具有更好的分类性能。
其他语种文摘 Hyperspectral image classification,which assigns each pixel to predefined land cover categories,is of crucial importance in various Earth science tasks such as environmental mapping and other related fields.In recent years,scholars have attempted to utilize deep learning frameworks for hyperspectral image classification and achieved satisfactory results.However,these methods still have certain deficiencies in extracting spectral features.This paper proposes a hierarchical self-attention network(HSAN)for hyperspectral image classification based on the self-attention mechanism.Firstly,a skip-layer self-attention module is constructed for feature learning,leveraging the self-attention mechanism of Transformer to capture contextual information and enhance the contribution of relevant information.Secondly,a hierarchical fusion method is designed to further alleviate the loss of relevant information during the feature learning process and enhance the interplay of features at different hierarchical levels.Experimental results on the Pavia University and Houston2013 datasets demonstrate that the proposed framework outperforms other state-of-the-art hyperspectral image classification frameworks.
来源 测绘学报 ,2023,52(7):1139-1147 【核心库】
DOI 10.11947/j.AGCS.2023.20220540
关键词 高光谱图像分类 ; Transformer ; 自注意力机制 ; 层级融合
地址

1. 中国科学院西安光学精密机械研究所, 中国科学院光谱成像技术重点实验室, 陕西, 西安, 710119  

2. 中国科学院大学, 北京, 100049  

3. 福州大学物理与信息工程学院, 福建, 福州, 350100

语种 中文
文献类型 研究性论文
ISSN 1001-1595
学科 测绘学
基金 国家自然科学基金 ;  国家自然科学基金国家杰出青年科学基金 ;  陕西省重点研发计划
文献收藏号 CSCD:7521098

参考文献 共 30 共2页

1.  Ahmad M. Hyperspectral image classification-traditional to deep models:a survey for future prospects. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2021 CSCD被引 1 次    
2.  Rasti B. Feature extraction for hyperspectral imagery:the evolution from shallow to deep:overview and toolbox. IEEE Geoscience and Remote Sensing Magazine,2020,8(4):60-88 CSCD被引 19 次    
3.  Xue Zhaohui. Low-Rank and Sparse Representation with Adaptive Neighborhood Regularization for Hyperspectral Image Classification. Journal of Geodesy and Geoinformation Science,2022,5(1):73-90 CSCD被引 3 次    
4.  左溪冰. 高光谱影像小样本分类的图卷积网络方法. 测绘学报,2021,50(10):1358-1369 CSCD被引 13 次    
5.  Ham J. Investigation of the random forest framework for classification of hyperspectral data. IEEE Transactions on Geoscience and Remote Sensing,2005,43(3):492-501 CSCD被引 47 次    
6.  Camps-Valls G. Kernel-based methods for hyperspectral image classification. IEEE Transactions on Geoscience and Remote Sensing,2005,43(6):1351-1362 CSCD被引 55 次    
7.  Ma Li. Local manifold learning-based K-nearest-neighbor for hyperspectral image classification. IEEE Transactions on Geoscience and Remote Sensing,2010,48(11):4099-4109 CSCD被引 28 次    
8.  Hu Wei. Deep convolutional neural networks for hyperspectral image classification. Journal of Sensors,2015 CSCD被引 5 次    
9.  Zhao Wenzhi. Spectral-spatial feature extraction for hyperspectral image classification:a dimension reduction and deep learning approach. IEEE Transactions on Geoscience and Remote Sensing,2016,54(8):4544-4554 CSCD被引 55 次    
10.  Wang Kexian. A Deep Double-Channel Dense Network for Hyperspectral Image Classification. Journal of Geodesy and Geoinformation Science,2021,4(4):46-62 CSCD被引 12 次    
11.  Liu Xun. StfNet: a two-stream convolutional neural network for spatiotemporal image fusion. IEEE Transactions on Geoscience and Remote Sensing,2019,57(9):6552-6564 CSCD被引 11 次    
12.  Wu Hao. Convolutional recurrent neural networks for hyperspectral data classification. Remote Sensing,2017,9(3):298 CSCD被引 8 次    
13.  Yu Shiqi. Convolutional neural networks for hyperspectral image classification. Neurocomputing,2017,219:88-98 CSCD被引 29 次    
14.  Gao Hongmin. Joint alternate small convolution and feature reuse for hyperspectral image classification. ISPRS International Journal of Geo-Information,2018,7(9):349 CSCD被引 1 次    
15.  Ran Lingyan. A hyperspectral image classification framework with spatial pixel pair features. Sensors,2017,17(10):2421 CSCD被引 3 次    
16.  Zhong Zilong. Spectral-spatial residual network for hyperspectral image classification:a 3Ddeep learning framework. IEEE Transactions on Geoscience and Remote Sensing,2017,56(2):847-858 CSCD被引 6 次    
17.  Li Simin. Adaptive spatialspectral feature learning for hyperspectral image classification. IEEE Access,2019,7:61534-61547 CSCD被引 2 次    
18.  Roy S K. Attentionbased adaptive spectral-spatial kernel ResNet for hyperspectral image classification. IEEE Transactions on Geoscience and Remote Sensing,2020,59(9):7831-7843 CSCD被引 3 次    
19.  Mou Lichao. Nonlocal graph convolutional networks for hyperspectral image classification. IEEE Transactions on Geoscience and Remote Sensing,2020,58(12):8246-8257 CSCD被引 15 次    
20.  Wan Sheng. Multiscale dynamic graph convolutional network for hyperspectral image classification. IEEE Transactions on Geoscience and Remote Sensing,2019,58(5):3162-3177 CSCD被引 21 次    
引证文献 4 篇

1 陆美 融合多尺度低秩表示与双向递归滤波的高光谱图像分类 遥感技术与应用,2024,39(2):393-404
CSCD被引 0 次

2 李永生 分波段Transformer特征提取在近红外光谱数据分类中的应用 激光与光电子学进展,2024,61(13):1330002
CSCD被引 0 次

显示所有4篇文献

论文科学数据集
PlumX Metrics
相关文献

 作者相关
 关键词相关
 参考文献相关

版权所有 ©2008 中国科学院文献情报中心 制作维护：中国科学院文献情报中心
地址：北京中关村北四环西路33号 邮政编码：100190 联系电话：(010)82627496 E-mail:cscd@mail.las.ac.cn 京ICP备05002861号-4 | 京公网安备11010802043238号