Dealing with Heterogeneous Google Earth Images on Building Area Detection Task

  • Cassio AlmeidaEmail author
  • William Fernandes
  • Simone Barbosa
  • Hélio Lopes
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11401)


This paper proposes the use of a convolutional neural network in the building detection task, using as data source the high resolution heterogeneous Google Earth RGB images. This is a challenging task not only because it is necessary to build a training dataset, but also because the learning model has to respond well to the diversity of images taking into account a relatively small training dataset. We use images from Brazil that have a very diverse geography due to their continental dimension. In this case, Google Earth images are very heterogeneous and represent a great challenge. We use an adapted U-net and another convolutional neural network classification method recently proposed in the literature for solving this particular task. The use of U-net shows promising results.


Convolution neural network Image segmentation Build detection 



We gratefully acknowledge the support of NVIDIA Corporation with the donation of the GPU used for this research. We would also like to thank Google Earth and their partners for the images utilized in this study. We also thanks CAPES and CNPq for their partial support.


  1. 1.
    Almeida, C.F.P., Fernandes, W.P.D., Barbosa, S.D.J., Lopes, H.: Using U-net to identify building areas on Google earth images from Brazil - annotated images (2017)Google Scholar
  2. 2.
    Cetin, M., Halici, U., Aytekin, Ö.: Building detection in satellite images by textural features and Adaboost. In: 2010 IAPR Workshop on PRRS, pp. 1–4. IEEE (2010)Google Scholar
  3. 3.
    Chen, L., Zhao, S., Han, W., Li, Y.: Building detection in an urban area using lidar data and quickbird imagery. IJRS 33(16), 5135–5148 (2012)Google Scholar
  4. 4.
    Cheng, G., Han, J.: A survey on object detection in optical remote sensing images. ISPRS 117, 11–28 (2016)CrossRefGoogle Scholar
  5. 5.
    Chollet, F., et al.: Keras (2015).
  6. 6.
    Cireşan, D., Meier, U., Schmidhuber, J.: Multi-column deep neural networks for image classification. In: ICPR, pp. 3642–3649 (2012)Google Scholar
  7. 7.
    Dornaika, F., Moujahid, A., El, Y., Ruichek, Y.: Building detection from orthophotos using a machine learning approach: an empirical study on image segmentation and descriptors. Expert Syst. Appl. 58, 130–142 (2016)CrossRefGoogle Scholar
  8. 8.
    Ghaffarian, S., Ghaffarian, S.: Automatic building detection based on Purposive FastICA (PFICA) algorithm using monocular high resolution Google Earth images. ISPRS 97, 152–159 (2014)CrossRefGoogle Scholar
  9. 9.
    Guo, J., Liang, L., Gong, P.: Removing shadows from google earth images. Int. J. Remote Sens. 31(6), 1379–1389 (2010)CrossRefGoogle Scholar
  10. 10.
    Guo, Y., Liu, Y., Oerlemans, A., Lao, S., Wu, S., Lew, M.S.: Deep learning for visual understanding: a review. Neurocomputing 187, 27–48 (2016)CrossRefGoogle Scholar
  11. 11.
    Guo, Z., Shao, X., Xu, Y., Miyazaki, H., Ohira, W., Shibasaki, R.: Identification of village building via google earth images and supervised machine learning methods. Remote Sens. 8(4), 271 (2016)CrossRefGoogle Scholar
  12. 12.
    Kampffmeyer, M., Salberg, A.B., Jenssen, R.: Semantic segmentation of small objects and modeling of uncertainty in urban remote sensing images using deep convolutional neural networks. In: CVPRW, pp. 1–9. IEEE (2016)Google Scholar
  13. 13.
    Karpatne, A., Jiang, Z., Vatsavai, R.R., Shekhar, S., Kumar, V.: Monitoring land-cover changes: a machine-learning perspective. IEEE GRSM 4(2), 8–21 (2016)Google Scholar
  14. 14.
    Längkvist, M., Kiselev, A., Alirezaie, M., Loutfi, A.: Classification and segmentation of satellite orthoimagery using CNN. Remote Sens. 8, 329 (2016)CrossRefGoogle Scholar
  15. 15.
    Liasis, G., Stavrou, S.: Building extraction in satellite images using active contours and colour features. IJRS 37(5), 1127–1153 (2016)Google Scholar
  16. 16.
    Long, J., Shelhamer, E., Darrell, T.: Fully convolutional networks for semantic segmentation. In: CVPR, 07–12 June 2015, pp. 3431–3440 (2015)Google Scholar
  17. 17.
    Mishra, A., Pandey, A., Baghel, A.S.: Building detection and extraction techniques: a review. In: Computing for Sustainable Global Development, pp. 3816–3821 (2016)Google Scholar
  18. 18.
    Quang, N.T., Thuy, N.T., Sang, D.V., Binh, H.T.T.: An efficient framework for pixel-wise building segmentation from aerial images. In: SoICT, pp. 282–287. ACM (2015)Google Scholar
  19. 19.
    Ronneberger, O., Fischer, P., Brox, T.: U-net: convolutional networks for biomedical image segmentation. In: Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F. (eds.) MICCAI 2015. LNCS, vol. 9351, pp. 234–241. Springer, Cham (2015). Scholar
  20. 20.
    Yang, Y., Newsam, S.: Bag-of-visual-words and spatial extensions for land-use classification. In: Proceedings of the 18th SIGSPATIAL International Conference on Advances in Geographic Information Systems, pp. 270–279. ACM (2010)Google Scholar
  21. 21.
    Zhang, L., Zhang, L., Kumar, V.: Deep learning for remote sensing data. IEEE Geosci. Remote. Sens. Mag. 4(2), 18 (2016)Google Scholar
  22. 22.
    Zhang, Q., Wang, Y., Liu, Q., Liu, X., Wang, W.: CNN based suburban building detection using monocular high resolution Google earth images. In: IGARSS, pp. 661–664. IEEE (2016)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Cassio Almeida
    • 1
    • 2
    Email author
  • William Fernandes
    • 1
  • Simone Barbosa
    • 1
  • Hélio Lopes
    • 1
  1. 1.Departamento de InformáticaPUC-RioRio de JaneiroBrazil
  2. 2.ENCE-IBGERio de JaneiroBrazil

Personalised recommendations