Wetland classification based on Sentinel-2 and 3D multisource domain self-attention model
- Vol. 28, Issue 1, Pages: 266-279(2024)
Published: 07 January 2024
DOI: 10.11834/jrs.20232246
扫 描 看 全 文
浏览全部资源
扫码关注微信
Published: 07 January 2024 ,
扫 描 看 全 文
楼桉君,贺智,肖曼,李心媛.2024.基于Sentinel-2和3D多源域自注意力模型的湿地分类.遥感学报,28(1): 266-279
Lou A J,He Z,Xiao M and Li X Y. 2024. Wetland classification based on Sentinel-2 and 3D multisource domain self-attention model. National Remote Sensing Bulletin, 28(1):266-279
准确的湿地分类可掌握湿地时空分异特征,在湿地研究中占据重要地位。针对现有基于小样本学习的湿地分类方法仅局限于利用目标域或单源域数据的问题,本文提出一种3D多源域自注意力小样本学习模型3D-MDAFSL(3D
M
ulti-source
D
omain self-
A
ttention
F
ew-
S
hot
L
earning)。首先,结合卷积和注意力机制的优势,设计基于自注意力机制和深度残差卷积的3D特征提取器;然后,采用对抗域自适应策略实现多源域特征对齐,在每个域分别进行小样本学习;最后,利用训练好的模型提取特征,并将特征输入至K近邻(K-nearest Neighbor)分类器以获取分类结果。结果表明,3D特征提取器相比无特征提取框架的湿地总体分类精度提升约6.79%;当使用多源域数据集时,3D-MDAFSL模型对中山市Sentinel-2湿地数据集的总体分类精度能达到93.52%,相比于现有算法有明显提升。本文所提出的3D-MDAFSL模型在湿地地物高精度提取和分类中有较好的应用价值。
Accurate wetland classification methods can quickly grasp the spatial-temporal variation characteristics of wetlands and play an important role in wetland research. Considering the limitation of the existing wetland classification method based on few-shot learning to the use of target or single-source domain dataset
this paper proposes a 3D multisource domain self-attention few-shot learning (3D-MDAFSL) model. First
combining the advantages of convolution and attention mechanism
a 3D feature extractor based on self-attention mechanism and deep residual convolution is designed. Then
the conditional adversarial domain adaptation strategy is used to achieve multisource domain distribution alignment
and few-shot learning is performed separately in each domain. Finally
the features extracted by the trained model are imputed to the K-nearest neighbor classifier to obtain classification results. Results show that compared with the framework without feature extraction
the 3D feature extractor improves the overall accuracy by approximately 6.79%. When using multisource domain datasets
the overall accuracy of the 3D-MDAFSL model for the Sentinel-2 wetland dataset in Zhongshan City can reach 93.52%
which is a significant improvement compared with the existing algorithms. The 3D-MDAFSL model proposed in this paper has good application value in the high-precision extraction and classification of wetland features.
遥感小样本湿地分类多源域自注意力
remote sensingfew-shotwetland classificationmulti-source domain adaptionself-attention
Bahdanau D, Cho K and Bengio Y. 2014. Neural machine translation by jointly learning to align and translate//Proceedings of the 3rd International Conference on Learning Representations. San Diego: [s.l.]
Cao J Z, Li Y W, Zhang K and Van Gool L. 2021. Video super-resolution transformer. arXiv:2106.06847 [DOI: 10.48550/arXiv.2106.06847http://dx.doi.org/10.48550/arXiv.2106.06847]
Cao Z Y, Li X R, Jiang J F and Zhao L Y. 2020. 3D convolutional siamese network for few-shot hyperspectral classification. Journal of Applied Remote Sensing, 14(4): 048504 [DOI: 10.1117/1.JRS.14.048504http://dx.doi.org/10.1117/1.JRS.14.048504]
Chen N X, Watanabe S, Villalba J, Żelasko P and Dehak N. 2021. Non-autoregressive transformer for speech recognition. IEEE Signal Processing Letters, 28: 121-125 [DOI: 10.1109/LSP.2020.3044547http://dx.doi.org/10.1109/LSP.2020.3044547]
Cho K, Van Merriënboer B, Gulcehre C, Bahdanau D, Bougares F, Schwenk H and Bengio Y. 2014. Learning phrase representations using RNN encoder-decoder for statistical machine translation//Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing. Doha: ACL [DOI: 10.3115/v1/D14-1179http://dx.doi.org/10.3115/v1/D14-1179]
Dai Z H, Liu H X, Le Q V and Tan M X. 2021. CoAtNet: marrying convolution and attention for all data sizes//Proceedings of the 35th Conference on Neural Information Processing Systems. [s.l.]: [s.n.]: 3965-3977
Deng C, Liu X L, Li C and Tao D C. 2018. Active multi-kernel domain adaptation for hyperspectral image classification. Pattern Recognition, 77: 306-315 [DOI: 10.1016/j.patcog.2017.10.007http://dx.doi.org/10.1016/j.patcog.2017.10.007]
Deng J, Dong W, Socher R, Li L J, Li K and Li F F. 2009. ImageNet: a large-scale hierarchical image database//Proceedings of 2009 IEEE Conference on Computer Vision and Pattern Recognition. Miami: IEEE [DOI: 10.1109/CVPR.2009.5206848http://dx.doi.org/10.1109/CVPR.2009.5206848]
Dosovitskiy A, Beyer L, Kolesnikov A, Weissenborn D, Zhai X H, Unterthiner T, Dehghani M, Minderer M, Heigold G and Gelly S, Uszkoreit J and Houlsby N. 2021. An image is worth 16x16 words: transformers for image recognition at scale//Proceedings of the 9th International Conference on Learning Representations. [s.l.]: OpenReview.net
Fu B L, Liu M, He H C, Lan F W, He X, Liu L L, Huang L K, Fan D L, Zhao M and Jia Z L. 2021. Comparison of optimized object-based rf-dt algorithm and segnet algorithm for classifying karst wetland vegetation communities using ultra-high spatial resolution uav data. International Journal of Applied Earth Observation and Geoinformation, 104: 102553 [DOI: 10.1016/j.jag.2021.102553http://dx.doi.org/10.1016/j.jag.2021.102553]
Gao K L, Liu B, Yu X C, Qin J C, Zhang P Q and Tan X. 2020. Deep relation network for hyperspectral image few-shot classification. Remote Sensing, 12(6): 923 [DOI: 10.3390/rs12060923http://dx.doi.org/10.3390/rs12060923]
Gong J Y and Ji S P. 2017. From photogrammetry to computer vision. Geomatics and Information Science of Wuhan University, 42(11): 1518-1522, 1615
龚健雅, 季顺平. 2017. 从摄影测量到计算机视觉. 武汉大学学报(信息科学版), 42(11): 1518-1522, 615 [DOI: 10.13203/j.whugis20170283http://dx.doi.org/10.13203/j.whugis20170283]
Gong R, Dai D X, Chen Y H, Li W and Van Gool L. 2021. mDALU: multi-source domain adaptation and label unification with partial datasets//Proceedings of 2021 IEEE/CVF International Conference on Computer Vision. Montreal: IEEE [DOI: 10.1109/ICCV48922.2021.00875http://dx.doi.org/10.1109/ICCV48922.2021.00875]
Gu Y X, Lan X, Fu B Y and Qin X L. 2023. Object detection algorithm for remote sensing images based on geometric adaptation and global perception. Journal of Computer Applications, 43(3): 916-922
顾勇翔, 蓝鑫, 伏博毅, 秦小林. 2023. 基于几何适应与全局感知的遥感图像目标检测算法. 计算机应用, 43(3): 916-922 [DOI: 10.11772/j.issn.1001-9081.2022010071http://dx.doi.org/10.11772/j.issn.1001-9081.2022010071]
Guo H J, Cai Y P, Yang Z F, Zhu Z C and Ouyang Y R. 2021. Dynamic simulation of coastal wetlands for Guangdong-Hong Kong-Macao Greater Bay area based on multi-temporal Landsat images and FLUS model. Ecological Indicators, 125: 107559 [DOI: 10.1016/j.ecolind.2021.107559http://dx.doi.org/10.1016/j.ecolind.2021.107559]
Gxokwe S, Dube T and Mazvimavi D. 2022. Leveraging Google Earth Engine platform to characterize and map small seasonal wetlands in the semi-arid environments of South Africa. Science of the Total Environment, 803: 150139 [DOI: 10.1016/j.scitotenv.2021.150139http://dx.doi.org/10.1016/j.scitotenv.2021.150139]
Hu Q, Woldt W, Neale C, Zhou Y Z, Drahota J, Varner D, Bishop A, LaGrange T, Zhang L G and Tang Z H. 2021. Utilizing unsupervised learning, multi-view imaging, and CNN-based attention facilitates cost-effective wetland mapping. Remote Sensing of Environment, 267: 112757 [DOI: 10.1016/j.rse.2021.112757http://dx.doi.org/10.1016/j.rse.2021.112757]
Jiao L C, Liang M M, Chen H, Yang S Y, Liu H Y and Cao X H. 2017. Deep fully convolutional network-based spatial distribution prediction for hyperspectral image classification. IEEE Transactions on Geoscience and Remote Sensing, 55(10): 5585-5599 [DOI: 10.1109/TGRS.2017.2710079http://dx.doi.org/10.1109/TGRS.2017.2710079]
Li Z K, Liu M, Chen Y S, Xu Y M, Li W and Du Q. 2022. Deep cross-domain few-shot learning for hyperspectral image classification. IEEE Transactions on Geoscience and Remote Sensing, 60: 5501618 [DOI: 10.1109/TGRS.2021.3057066http://dx.doi.org/10.1109/TGRS.2021.3057066]
Liu B, Yu X C, Yu A Z, Zhang P Q, Wan G and Wang R R. 2019. Deep few-shot learning for hyperspectral image classification. IEEE Transactions on Geoscience and Remote Sensing, 57(4): 2290-2304 [DOI: 10.1109/TGRS.2018.2872830http://dx.doi.org/10.1109/TGRS.2018.2872830]
Liu W. 2020. Domain Adaptation for the High Resolution Remote Sensing Image Classification. Harbin: Harbin Institute of Technology
刘玮. 2020. 面向高分辨率遥感图像分类的域适应算法研究. 哈尔滨: 哈尔滨工业大学 [DOI: 10.27061/d.cnki.ghgdu.2020.004928http://dx.doi.org/10.27061/d.cnki.ghgdu.2020.004928]
Mei S H, Ji J Y, Hou J H, Li X and Du Q. 2017. Learning sensor-specific spatial-spectral features of hyperspectral images via convolutional neural networks. IEEE Transactions on Geoscience and Remote Sensing, 55(8): 4520-4533 [DOI: 10.1109/TGRS.2017.2693346http://dx.doi.org/10.1109/TGRS.2017.2693346]
Meng X R. 2019. Study on Remote Sensing Finer Classification of Freshwater Wetland Based on Deep Learning. Harbin: Harbin Normal University
孟祥锐. 2019. 基于深度学习的淡水湿地遥感精细分类研究. 哈尔滨: 哈尔滨师范大学 [DOI: 10.27064/d.cnki.ghasu.2019.000087http://dx.doi.org/10.27064/d.cnki.ghasu.2019.000087]
Mo L J, Cao Y, Hu Y M, Liu M and Xiao D. 2012. Object-oriented classification for satellite remote sensing of wetlands: a case study in southern Hangzhou bay area. Wetland Science, 10(2): 206-213
莫利江, 曹宇, 胡远满, 刘淼, 夏栋. 2012. 面向对象的湿地景观遥感分类——以杭州湾南岸地区为例. 湿地科学, 10(2): 206-213 [DOI: 10.13248/j.cnki.wetlandsci.2012.02.019http://dx.doi.org/10.13248/j.cnki.wetlandsci.2012.02.019]
Ning X G, Chang W T, Wang H, Zhang H C and Zhu Q D. 2022. Extraction of marsh wetland in Heilongjiang Basin based on GEE and multi-source remote sensing data. National Remote Sensing Bulletin, 26(2): 386-396
宁晓刚, 常文涛, 王浩, 张翰超, 朱乾德. 2022. 联合GEE与多源遥感数据的黑龙江流域沼泽湿地信息提取. 遥感学报, 26(2): 386-396 [DOI: 10.11834/jrs.20200033http://dx.doi.org/10.11834/jrs.20200033]
Othman E, Bazi Y, Melgani F, Alhichri H, Alajlan N and Zuair M. 2017. Domain adaptation network for cross-scene classification. IEEE Transactions on Geoscience and Remote Sensing, 55(8): 4441-4456 [DOI: 10.1109/TGRS.2017.2692281http://dx.doi.org/10.1109/TGRS.2017.2692281]
Rao M B, Tang P and Zhang Z. 2019. Spatial–spectral relation network for hyperspectral image classification with limited training samples. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 12(12): 5086-5100 [DOI: 10.1109/JSTARS.2019.2957047http://dx.doi.org/10.1109/JSTARS.2019.2957047]
Shen J, Qu Y R, Zhang W N and Yu Y. 2018. Wasserstein distance guided representation learning for domain adaptation//Proceedings of the Thirty-Second AAAI Conference on Artificial Intelligence and Thirtieth Innovative Applications of Artificial Intelligence Conference and Eighth AAAI Symposium on Educational Advances in Artificial Intelligence. New Orleans: AAAI Press
Simonyan K and Zisserman A. 2015. Very deep convolutional networks for large-scale image recognition//Proceedings of the 3rd International Conference on Learning Representations. San Diego: [s.n.]
Ståhl N and Weimann L. 2022. Identifying wetland areas in historical maps using deep convolutional neural networks. Ecological Informatics, 68: 101557 [DOI: 10.1016/j.ecoinf.2022.101557http://dx.doi.org/10.1016/j.ecoinf.2022.101557]
Sutskever I, Vinyals O and Le Q V. 2014. Sequence to sequence learning with neural networks//Proceedings of the 27th International Conference on Neural Information Processing Systems. Montreal: MIT Press
Tuia D, Persello C and Bruzzone L. 2016. Domain adaptation for the classification of remote sensing data: An overview of recent advances. IEEE Geoscience and Remote Sensing Magazine, 4(2): 41-57 [DOI: 10.1109/MGRS.2016.2548504http://dx.doi.org/10.1109/MGRS.2016.2548504]
Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez A N, Kaiser Ł and Polosukhin I. 2017. Attention is all you need//Proceedings of the 31st International Conference on Neural Information Processing Systems. Long Beach: Curran Associates Inc
Wang G L, Fan B, Xiang S M and Pan C H. 2017. Aggregating rich hierarchical features for scene classification in remote sensing imagery. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 10(9): 4104-4115 [DOI: 10.1109/JSTARS.2017.2705419http://dx.doi.org/10.1109/JSTARS.2017.2705419]
Wang Z M, Du B, Shi Q and Tu W P. 2019. Domain adaptation with discriminative distribution and manifold embedding for hyperspectral image classification. IEEE Geoscience and Remote Sensing Letters, 16(7): 1155-1159 [DOI: 10.1109/LGRS.2018.2889967http://dx.doi.org/10.1109/LGRS.2018.2889967]
Yang H L and Crawford M M. 2016. Domain adaptation with preservation of manifold geometry for hyperspectral image classification. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 9(2): 543-555 [DOI: 10.1109/JSTARS.2015.2449738http://dx.doi.org/10.1109/JSTARS.2015.2449738]
Yang J X, Zhao Y Q and Chan J C W. 2017. Learning and transferring deep joint spectral–spatial features for hyperspectral classification. IEEE Transactions on Geoscience Remote Sensing, 55(8): 4729-4742 [DOI: 10.1109/TGRS.2017.2698503http://dx.doi.org/10.1109/TGRS.2017.2698503]
Yang L X, Zhang R Y, Li L D and Xie X H. 2021. SimAM: a simple, parameter-free attention module for convolutional neural networks//Proceedings of the 38th International Conference on Machine Learning. [s.l.]: PMLR
Yu S Q, Jia S, and Xu C Y. 2017. Convolutional neural networks for hyperspectral image classification. Neurocomputing, 219: 88-98 [DOI: 10.1016/j.neucom.2016.09.010http://dx.doi.org/10.1016/j.neucom.2016.09.010]
Zhang K Y. 2019. Remote Sensing Mapping and Evolution Analysis of Coastal Beach Based on Time Series Image Analysis of Google Earth Engine Platform. Shenzhen: Shenzhen University
张康永. 2019. 基于Google Earth Engine平台时序影像分析的滨海滩涂遥感制图及演变分析. 深圳: 深圳大学
Zhao S C, Wang G Z, Zhang S H, Gu Y, Li Y X, Song Z C, Xu P F, Hu R B, Chai H and Keutzer K. 2020. Multi-source distilling domain adaptation//Proceedings of the Thirty-Fourth AAAI Conference on Artificial Intelligencem, AAAI 2020, The Thirty-Second Innovative Applications of Artificial Intelligence Conference, IAAI 2020, The Tenth AAAI Symposium on Educational Advances in Artificial Intelligence. New York: AAAI Press [DOI: 10.1609/aaai.v34i07.6997http://dx.doi.org/10.1609/aaai.v34i07.6997]
相关文章
相关作者
相关机构