Advertisement

Chinese Text Detection Using Deep Learning Model and Synthetic Data

  • Wei-wei Gao
  • Jun Zhang
  • Peng Chen
  • Bing Wang
  • Yi Xia
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10954)

Abstract

Detection of text in natural scene images is very challenging, and it is not completely solved. In this work we propose a fast and reliable algorithm to generate synthetic data of Chinese characters in images. The proposed algorithm make the text content cover the background in a natural way. To validate the proposed method effective, another dataset are generated by ordinary fusion method. Two dataset are used to train Faster-RCNN network. And the experimental result shows that the dataset are generated by proposed method achieve a better performance of detection than the normal way.

Keywords

Synthetic data Text detection Faster R-CNN 

Notes

Acknowledge

This work is supported by Anhui Provincial Natural Science Foundation (grant number 1608085MF136).

References

  1. 1.
    Ren, S., He, K., Girshick, R., et al.: Faster R-CNN: towards real-time object detection with region proposal networks. In: International Conference on Neural Information Processing Systems, pp. 91–99. MIT Press (2015)Google Scholar
  2. 2.
    Geng, Y., Liang, R.Z., Li, W., et al.: Learning convolutional neural network to maximize Pos@Top performance measure (2016)Google Scholar
  3. 3.
    Geng, Y., et al.: A novel image tag completion method based on convolutional neural transformation. In: Lintas, A., Rovetta, S., Verschure, P.F.M.J., Villa, A.E.P. (eds.) ICANN 2017. LNCS, vol. 10614, pp. 539–546. Springer, Cham (2017).  https://doi.org/10.1007/978-3-319-68612-7_61CrossRefGoogle Scholar
  4. 4.
    Zhang, G., et al.: Learning convolutional ranking-score function by query preference regularization. In: Yin, H., et al. (eds.) IDEAL 2017. LNCS, vol. 10585, pp. 1–8. Springer, Cham (2017).  https://doi.org/10.1007/978-3-319-68935-7_1CrossRefGoogle Scholar
  5. 5.
    Gupta, A., Vedaldi, A., Zisserman, A.: Synthetic data for text localization in natural images, 2315–2324 (2016)Google Scholar
  6. 6.
    Wang, T., Wu, D.J., Coates, A., Ng, A.Y.: End-to-end text recognition with convolutional neural networks. In: Proceedings ICPR, pp. 3304–3308 (2012)Google Scholar
  7. 7.
    Jaderberg, M., Simonyan, K., Vedaldi, A., Zisserman, A.: Synthetic data and artificial neural networks for natural scene text recognition. In: Workshop on Deep Learning, NIPS (2014)Google Scholar
  8. 8.
    Dosovitskiy, A., Fischery, P., Ilg, E., et al.: FlowNet: learning optical flow with convolutional networks. In: IEEE International Conference on Computer Vision, pp. 2758–2766. IEEE Computer Society (2015)Google Scholar
  9. 9.
    Dosovitskiy, A., Brox, T.: Inverting visual representations with convolutional networks, pp. 4829–4837 (2015)Google Scholar
  10. 10.
    Yildirim, I., Kulkarni, T., Freiwald, W., et al.: Efficient analysis-by-synthesis in vision: a computational framework, behavioral tests, and comparison with neural representations. In: Conference of the Cognitive Science Society (2015)Google Scholar
  11. 11.
    Jaderberg, M., Vedaldi, A., Zisserman, A.: Deep features for text spotting. In: Fleet, D., Pajdla, T., Schiele, B., Tuytelaars, T. (eds.) ECCV 2014. LNCS, vol. 8692, pp. 512–528. Springer, Cham (2014).  https://doi.org/10.1007/978-3-319-10593-2_34CrossRefGoogle Scholar
  12. 12.
    Ozuysal, O.M., Fua, P., Lepetit, V.: Fast keypoint recognition in ten lines of code. In: IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp. 1–8. DBLP (2007)Google Scholar
  13. 13.
    Wang, K., Babenko, B., Belongie, S.: End-to-end scene text recognition. In: IEEE International Conference on Computer Vision, pp. 1457–1464. IEEE (2012)Google Scholar
  14. 14.
    Alsharif, O., Pineau, J.: End-to-end text recognition with hybrid HMM maxout model. Comput. Sci. (2013)Google Scholar
  15. 15.
    Bissacco, A., Cummins, M., Netzer, Y., et al.: PhotoOCR: reading text in uncontrolled conditions. In: IEEE International Conference on Computer Vision, pp. 785–792. IEEE (2014)Google Scholar
  16. 16.
    Szegedy, C., Toshev, A., Erhan, D.: Deep neural networks for object detection. Adv. Neural. Inf. Process. Syst. 26, 2553–2561 (2013)Google Scholar
  17. 17.
    Sermanet, P., Eigen, D., Zhang, X., Mathieu, M., Fergus, R., LeCun, Y.: OverFeat: integrated recognition, localization and detection using convolutional networks. In: ICLR (2014)Google Scholar
  18. 18.
    Erhan, D., Szegedy, C., Toshev, A., et al.: Scalable object detection using deep neural networks. 3(4), 2155–2162 (2013)Google Scholar
  19. 19.
    Szegedy, C., Reed, S., Erhan, D., et al.: Scalable, high-quality object detection. Comput. Sci. (2014)Google Scholar
  20. 20.
    Girshick, R., Donahue, J., Darrell, T., Malik, J.: Rich feature hierarchies for accurate object detection and semantic segmentation. In: CVPR (2014)Google Scholar
  21. 21.
    Arbelaez, P., Maire, M., Fowlkes, C., Malik, J.: Contour detection and hierarchical image segmentation. IEEE PAMI 33, 898–916 (2011)CrossRefGoogle Scholar
  22. 22.
    Liu, C.S., Lin, G.: Deep convolutional neural fields for depth estimation from a single image. In: Proceedings CVPR (2015)Google Scholar
  23. 23.
    Fischler, M.A., Bolles, R.C.: Random sample consensus: A paradigm for model fitting with applications to image analysis and automated cartography. Comm. ACM 24(6), 381–395 (1981)MathSciNetCrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  • Wei-wei Gao
    • 1
  • Jun Zhang
    • 1
  • Peng Chen
    • 2
  • Bing Wang
    • 3
  • Yi Xia
    • 1
  1. 1.School of Electrical Engineering and AutomationAnhui UniversityHefeiChina
  2. 2.Institute of Health SciencesAnhui UniversityHefeiChina
  3. 3.School of Electrical and Information EngineeringAnhui University of TechnologyMa AnshanChina

Personalised recommendations