WetlandNet: Semantic Segmentation for Remote Sensing Images of Coastal Wetlands via Improved UNet with Deconvolution

  • Binge CuiEmail author
  • Yonghui Zhang
  • Xinhui Li
  • Jing Wu
  • Yan Lu
Conference paper
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 1107)


In order to keep the clear boundary contour for segmented object in semantic segmentation of remote sensing image of coastal wetlands, a deep learning semantic segmentation model called WetlandNet is proposed by improving UNet. The model takes encoder-decoder as its basic structure, uses depthwise separable convolution instead of regular convolution to reduce the model parameters, uses deconvolution to extract boundary contour features of the object, and connects these features to upsampled feature maps by jump connection. This paper takes the remote sensing image of the Yellow River Estuary wetland in Kenli District, Dongying City, Shandong Province, China, taken by GF-2 satellite as an example. The experimental results show that compared with advanced semantic segmentation models UNet, PSPNet and DeepLabV3 in deep learning, the proposed model achieves more accurate segmentation results, which improves the OA by more than 5%, Kappa by more than 0.07, and the F1 scores of five of the six classes are higher than those of contrast models, while the number of parameters is only 1/36 of UNet, 1/42 of PSPNet and 1/51 of DeepLabV3.


Coastal wetland Remote sensing image Deep learning Semantic segmentation UNet 


  1. 1.
    Hu, Y., et al.: Hyperspectral coastal wetland classification based on a multiobject convolutional neural network model and decision fusion. IEEE Geosci. Remote Sens. Lett. PP(99), 1–5 (2019)Google Scholar
  2. 2.
    Woodward, R.T., Wui, Y.S.: The economic value of wetland services: a meta-analysis. Ecol. Econ. 37(2), 257–270 (2001)CrossRefGoogle Scholar
  3. 3.
    Yang, Y.: Main characteristics, progress and prospect of international wetland science research. Progr. Geogr. 21(2), 111–120 (2002)Google Scholar
  4. 4.
    Hu, Y., et al.: Deep learning classification of coastal wetland hyperspectral image combined spectra and texture features: a case study of Huanghe (Yellow) River Estuary wetland (2019)Google Scholar
  5. 5.
    Long, J., et al.: Fully convolutional networks for semantic segmentation. IEEE Trans. Pattern Anal. Mach. Intell. 39(4), 640–651 (2014)Google Scholar
  6. 6.
    Ronneberger, O., Fischer, P., Brox, T.: U-net: convolutional networks for biomedical image segmentation (2015)Google Scholar
  7. 7.
    Zhao, H., et al.: Pyramid scene parsing network (2016)Google Scholar
  8. 8.
    Chen, L.C., et al.: Rethinking atrous convolution for semantic image segmentation (2017)Google Scholar
  9. 9.
    Cho, K., et al.: Learning phrase representations using RNN encoder-decoder for statistical machine translation (2014)Google Scholar
  10. 10.
    Krishnan, D., Tay, T., Fergus, R.: Blind deconvolution using a normalized sparsity measure. In: Proceedings of the Computer Vision and Pattern Recognition, CVPR (2011)Google Scholar
  11. 11.
    Kaiser, L., Gomez, A.N., Chollet, F.: Depthwise separable convolutions for neural machine translation (2017)Google Scholar
  12. 12.
    Simonyan, K., Zisserman, A.J.C.S.: Very deep convolutional networks for large-scale image recognition (2014)Google Scholar
  13. 13.
    He, K., et al.: Deep residual learning for image recognition (2015)Google Scholar
  14. 14.
    Paszke, A., et al.: ENet: a deep neural network architecture for real-time semantic segmentation (2016)Google Scholar
  15. 15.
    Noh, H., Hong, S., Han, B.: Learning deconvolution network for semantic segmentation. In: IEEE International Conference on Computer Vision (2015)Google Scholar
  16. 16.
    Wu, C., et al.: A compact DNN: approaching googlenet-level accuracy of classification and domain adaptation (2017)Google Scholar
  17. 17.
    Howard, A.G., et al.: MobileNets: efficient convolutional neural networks for mobile vision applications (2017)Google Scholar
  18. 18.
    Zhang, X., et al., ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices. 2017Google Scholar
  19. 19.
    Chollet, F., Xception: deep learning with depthwise separable convolutions (2016)Google Scholar
  20. 20.
    Gould, S., et al.: On differentiating parameterized argmin and argmax problems with application to bi-level optimization (2016)Google Scholar
  21. 21.
    Srivastava, N., et al.: Dropout: a simple way to prevent neural networks from overfitting. J. Mach. Learn. Res. 15(1), 1929–1958 (2014)MathSciNetzbMATHGoogle Scholar
  22. 22.
    Kopf, J., et al.: Joint bilateral upsampling (2007)Google Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2020

Authors and Affiliations

  • Binge Cui
    • 1
    Email author
  • Yonghui Zhang
    • 1
  • Xinhui Li
    • 1
  • Jing Wu
    • 1
  • Yan Lu
    • 1
  1. 1.College of Computer Science and EngineeringShandong University of Science and TechnologyQingdaoChina

Personalised recommendations