DOI QR코드

DOI QR Code

Performance Analysis of Detecting buried pipelines in GPR images using Faster R-CNN

Faster R-CNN을 활용한 GPR 영상에서의 지하배관 위치추적 성능분석

  • Ko, Hyoung-Yong (Department of Computer Science, Kyonggi University) ;
  • Kim, Nam-gi (Department of Computer Science, Kyonggi University)
  • 고형용 (경기대학교 컴퓨터과학과) ;
  • 김남기 (경기대학교 컴퓨터과학과)
  • Received : 2019.03.15
  • Accepted : 2019.05.20
  • Published : 2019.05.28

Abstract

Various pipes are buried in the city as needed, such as water pipes, gas pipes and hydrogen pipes. As the time passes, buried pipes becomes aged due to crack, etc. these pipes has the risk of accidents such as explosion and leakage. To prevent the risks, many pipes are repaired or replaced, but the location of the pipes can also be changed. Failure to identify the location of the altered pipe may cause an accident by touching the pipe. In this paper, we propose a method to detect buried pipes by gathering the GPR images by using GPR and Learning with Faster R-CNN. Then experiments was carried out by raw data sets and data sets augmentation applied to increase the amount of images.

JKOHBZ_2019_v9n5_21_f0001.png 이미지

Fig. 1. GPR Exploration path

JKOHBZ_2019_v9n5_21_f0002.png 이미지

Fig. 2. GPR Sample Image

JKOHBZ_2019_v9n5_21_f0003.png 이미지

Fig. 3. Faster R-CNN Process

JKOHBZ_2019_v9n5_21_f0004.png 이미지

Fig. 4. Faster R-CNN 13-Layer Structure

JKOHBZ_2019_v9n5_21_f0005.png 이미지

Fig. 5. Faster R-CNN Learning Process

JKOHBZ_2019_v9n5_21_f0006.png 이미지

Fig. 6. Data set A

JKOHBZ_2019_v9n5_21_f0007.png 이미지

Fig. 7. Data set B

JKOHBZ_2019_v9n5_21_f0008.png 이미지

Fig. 8. Data set C

JKOHBZ_2019_v9n5_21_f0009.png 이미지

Fig. 9. Data set A verification image example

JKOHBZ_2019_v9n5_21_f0010.png 이미지

Fig. 10. Data set B verification image example

JKOHBZ_2019_v9n5_21_f0011.png 이미지

Fig. 11. Data set C verification image example

JKOHBZ_2019_v9n5_21_f0012.png 이미지

Fig. 12. Graph result with augmentation applied

Table 1. Experiment Environment

JKOHBZ_2019_v9n5_21_t0001.png 이미지

Table 2. Experiment result symbol meaning

JKOHBZ_2019_v9n5_21_t0002.png 이미지

Table 3. Experiment result

JKOHBZ_2019_v9n5_21_t0003.png 이미지

Table 4. Result with augmentation applied

JKOHBZ_2019_v9n5_21_t0004.png 이미지

Acknowledgement

Supported by : Kyonggi University, National Research Foundation of Korea

References

  1. D. Daniels. (2004). Ground penetrating rader. The Institution of Engineering and Technology Publishing.
  2. D. Y. Lee. (2018). Analysis of sewer pipe defect and ground subsidence risk by using CCTV and GPR monitering results. Journal of Korean Geosynthetics Society, 17(3), 47-55. DOI : 10.12814/jkgss.2018.17.3.047 https://doi.org/10.12814/jkgss.2018.17.3.047
  3. M. G. Huici. (2013). Accurate ground penetrating radar numerical modeling for automatic detection and recognition of antipersonnel landmines. Doctoral dissertation. Rheinischen Friedrich Wilhelms Universitat Bonn, Bonn.
  4. Q. Dou, L. Wei, D. R. Magee & A. G. Cohn. (2017). Real-time hyperbola recognition and fitting in GPR data. IEEE Transactions on Geoscience and Remote Sensing, 55(1), 51-62. DOI : 10.1109/TGRS.2016.2592679 https://doi.org/10.1109/TGRS.2016.2592679
  5. M. C. Lee, H. C. Shin & J. H. Yoon. (2011). A study on the ground following and location marking method for mine detection system. Journal of the Korea Institute of Military Science and Technology, 14(6), 1002-1008. DOI : 10.9766/KIMST.2011.14.6.1002 https://doi.org/10.9766/KIMST.2011.14.6.1002
  6. R. Girshick, J. Donahue, T. Darrell & J. Malik. (2014). Rich feature hierarchies for accurate object detection and semantic segmentation. 2014 IEEE Conference on Computer Vision and Pattern Recognition (pp. 580-587). DOI : 10.1109/CVPR.2014.81 https://doi.org/10.1109/CVPR.2014.81
  7. R. Girshick. (2015, Sep). Fast R-CNN. 2015 International Conference on Computer Vision (pp. 1440-1448). DOI : 10.1109/ICCV.2015.169 https://doi.org/10.1109/ICCV.2015.169
  8. S. Ren, K. He, R. Girshick & J. Sun. (2017). Faster R-CNN: Towards real-time object detection with region proposal networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(6), 1137-1149. DOI : 10.1109/TPAMI.2016.2577031 https://doi.org/10.1109/TPAMI.2016.2577031
  9. S. J. Shin, J. K. Park, Y. Y. Kim, J. H. Jang & I. C. Moon. (2017). Analsys of the effect of convolutional neural network model on training and prediction in Faster R-CNN. Journal of Korean Institute of Industrial Engineers, 1-5.
  10. K. Simonyan & A. Zisserman. (2015, Apr). Very deep convolutional networks for large-scale image recognition. Proceedings of International Conference on Learning Representations (pp. 1-14).
  11. C. J. Kim, S. K. Ko, B. E. Kim, H. D. Jung & C. S. Cho. (2018). Object detection based on data augmentation using image rotation. Journal of Institute of Electronics and Information Engineers, 545-548.