WHU-RuR: Conjunto de dados de referência para extração de estradas rurais (terras agrícolas de alto padrão) a partir de imagens de satélite de alta resolução

  • role: First author第一作者
  • Affiliation:

    School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430079, China

    College of Geomatics, Xi'an University of Science and Technology, Xi'an 710054, China

  • Email:wangningjing@stu.xust.edu.cn
  • Introduction:王宁静,研究方向为高分辨率遥感图像智能解译。E-mail: wangningjing@stu.xust.edu.cn
WANG Ningjing12,  
  • role: Corresponding author通信作者
  • Affiliation:

    School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430079, China

    Institute of Aerospace Science and Technology, Wuhan University, Wuhan 430079, China

  • Email:wangxinyu@whu.edu.cn
  • Introduction:王心宇,研究方向为遥感信息智能处理与应用。E-mail: wangxinyu@whu.edu.cn
WANG Xinyu14*,  
  • Affiliation:

    State Key Laboratory of Information Engineering in Surveying, Wuhan University, Wuhan 430079, China

PAN Yang3,  
  • Affiliation:

    Department of Land Surveying and Geoinformatics, The Hong Kong Polytechnic University, Hong Kong 999077, China

LU Xiaoyan5,  
  • Affiliation:

    College of Geomatics, Xi'an University of Science and Technology, Xi'an 710054, China

YAO Wanqiang2,  
  • Affiliation:

    State Key Laboratory of Information Engineering in Surveying, Wuhan University, Wuhan 430079, China

ZHONG Yanfei3,  
  • Affiliation:

    School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430079, China

    Institute of Aerospace Science and Technology, Wuhan University, Wuhan 430079, China

GONG Jianya14

Resumo

A construção de terras agrícolas de alto padrão é uma medida fundamental para aumentar a capacidade de produção de alimentos na China e garantir a segurança alimentar. Entre elas, as estradas rurais (terras agrícolas de alto padrão), como caminhos de campo e estradas de produção, são um dos principais conteúdos da infraestrutura das terras agrícolas de alto padrão. A forma de utilizar imagens de satélite de alta resolução para extrair rapidamente e com precisão as estradas rurais é de grande importância para o monitoramento e supervisão da construção de terras agrícolas de alto padrão. Nos últimos anos, os métodos de aprendizado profundo para extração de estradas urbanas e os conjuntos de dados de referência se desenvolveram rapidamente, mas devido à grande diferença na distribuição das estradas urbanas e rurais, os modelos treinados para extrair estradas urbanas são difíceis de aplicar diretamente à extração de estradas rurais. Este artigo, baseado em dados de imagens de satélite de alta resolução, construiu sistematicamente o primeiro conjunto de dados de referência para extração de estradas rurais (terras agrícolas de alto padrão) de imagens de satélite de alta resolução (WHU-RuR). Este conjunto de dados é atualmente o maior conjunto de dados open source em termos de cobertura de terras agrícolas, o maior número de categorias de estradas rurais, o fundo rural mais complexo e o maior volume de dados, com as seguintes características: cobre sete províncias típicas da China (Hubei, Hunan, Shaanxi, Sichuan, Anhui, Henan e Hebei) no norte, centro, leste, noroeste e sudoeste; inclui ao todo 25922 pares de imagens de satélite de alta resolução e amostras de estradas (resolução espacial de 0,3 m, tamanho espacial de 1024×1024 pixels) em conjuntos de treinamento e teste. Ao mesmo tempo, para verificar a usabilidade do conjunto de dados WHU-RuR, este artigo testou e analisou de forma abrangente a eficácia dos métodos avançados de aprendizado profundo na tarefa de extração de estradas rurais. Os experimentos mostram que o conjunto de dados WHU-RuR basicamente atende às necessidades de extração de estradas rurais e possui grande potencial de aplicação no campo de monitoramento e supervisão das terras agrícolas de alto padrão. Link do conjunto de dados: https://doi.org/10.57760/sciencedb.09181.

Palavras-chave

extração de estradas rurais; terras agrícolas de alto padrão; imagens de satélite de alta resolução; conjunto de dados de referência para sensoriamento remoto; aprendizado profundo

References

  1. 1.
    Chaurasia A and Culurciello E. 2017. LinkNet: exploiting encoder representations for efficient semantic segmentation//2017 IEEE Visual Communications and Image Processing (VCIP). St. Petersburg: IEEE: 1-4
  2. 2.
    Chen L C, Zhu Y K, Papandreou G, Schroff F and Adam H. 2018. Encoder-decoder with atrous separable convolution for semantic image segmentation. arXiv:1802.02611
  3. 3.
    Demir I, Koperski K, Lindenbaum D, Pang G, Huang J, Basu S, Hughes F, Tuia D and Raskar R. 2018. DeepGlobe 2018: a challenge to parse the earth through satellite images//Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Salt Lake City: IEEE: 172-181
  4. 4.
    Geng K, Sun X, Yan Z Y, Diao W H and Gao X. 2020. Topological space knowledge distillation for compact road extraction in optical remote sensing images. Remote Sensing, 12(19): 3175
  5. 5.
    Lu X Y, Zhong Y F, Zheng Z and Zhang L P. 2021. GAMSNet: globally aware road detection network with multi-scale residual learning. ISPRS Journal of Photogrammetry and Remote Sensing, 175: 340-352
  6. 6.
    Mnih V. 2013. Machine Learning for Aerial Image Labeling. Toronto: University of Toronto
  7. 7.
    National People’s Congress Standing Committee of the People’s Republic of China. 2019. Land Administration Law of the People’s Republic of China
  8. 8.
    Pan Y, Wang X Y, Zhang L P and Zhong Y F. 2023. E2EVAP: end-to-end vectorization of smallholder agricultural parcel boundaries from high-resolution remote sensing imagery. ISPRS Journal of Photogrammetry and Remote Sensing, 203: 246-264
  9. 9.
    Ronneberger O, Fischer P and Brox T. 2015. U-net: convolutional networks for biomedical image segmentation//18th International Conference on Medical Image Computing and Computer-Assisted Intervention–MICCAI 2015. Munich: Springer: 234-241
  10. 10.
    State Council of the People’s Republic of China. 2011. Regulations on the Protection of Prime Farmland (中华人民共和国国务院. 2011. 基本农田保护条例)
  11. 11.
    Van Etten A, Lindenbaum D and Bacastow T M. 2019. SpaceNet: a remote sensing dataset and challenge series. arXiv preprint arXiv: 1807.01232
  12. 12.
    Wang J J, Zheng Z, Ma A L, Lu X Y and Zhong Y F. 2022. LoveDA: a remote sensing land-cover dataset for domain adaptive semantic segmentation. arXiv preprint arXiv: 2110.08733
  13. 13.
    Wei Y N, Wang Z L and Xu M. 2017. Road structure refined CNN for road extraction in aerial image. IEEE Geoscience and Remote Sensing Letters, 14(5): 709-713
  14. 14.
    Wu Q Q, Wang S, Wang B and Wu Y L. 2022. Road extraction method of high-resolution remote sensing image on the basis of the spatial information perception semantic segmentation model. National Remote Sensing Bulletin, 26(9): 1872-1885
  15. 15.
    Wu T, Xiang L G and Gong J Y. 2016. Updating road networks by local renewal from GPS trajectories. ISPRS International Journal of Geo-Information, 5(9): 163
  16. 16.
    Yang D F, Zhao J W, Li Y F, Xiao P and Yang J L. 2023. Iterative nearest edge algorithm for aerial image road dataset preparation. Acta Geodaetica et Cartographica Sinica, 52(1): 82-92
  17. 17.
    Yang Z G, Zhou D X, Yang Y, Zhang J P and Chen Z H. 2023. Road extraction from satellite imagery by road context and full-stage feature. IEEE Geoscience and Remote Sensing Letters, 20: 8000405
  18. 18.
    Zhang Z X, Liu Q J and Wang Y H. 2018. Road extraction by deep residual U-net. IEEE Geoscience and Remote Sensing Letters, 15(5): 749-753
  19. 19.
    Zhong Y F, Hu X, Luo C, Wang X Y, Zhao J and Zhang L P. 2020. WHU-Hi: UAV-borne hyperspectral with high spatial resolution (H2) benchmark datasets and classifier for precise crop identification based on deep convolutional neural network with CRF. Remote Sensing of Environment, 250: 112012
  20. 20.
    Zhong Y F, Yan B W, Yi J J, Yang R Y, Xu M Z, Su Y, Zheng Z D and Zhang L P. 2023. Global urban high-resolution land-use mapping: from benchmarks to multi-megacity applications. Remote Sensing of Environment, 298: 113758
  21. 21.
    Zhou L C, Zhang C and Wu M. 2018. D-LinkNet: LinkNet with pretrained encoder and dilated convolution for high resolution satellite imagery road extraction//Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Salt Lake City: IEEE: 182-186
  22. 22.
    Zhou M T, Sui H, Chen S X, Wang J D and Chen X. 2020. BT-RoadNet: a boundary and topologically-aware neural network for road extraction from high-resolution remote sensing imagery. ISPRS Journal of Photogrammetry and Remote Sensing, 168: 288-306
  23. 23.
    Zhu Q Q, Guo X, Deng W H, Shi S N, Guan Q F, Zhong Y F, Zhang L P and Li D R. 2022. Land-use/land-cover change detection based on a Siamese global learning framework for high spatial resolution remote sensing imagery. ISPRS Journal of Photogrammetry and Remote Sensing, 184: 63-78

Leia o texto completo

The above content is generated by Large Model Translation. The translated content is for reference only. We do not assume any commercial or legal responsibilty for any consequences arising from the use of our website