DOI QR코드

DOI QR Code

Development of Fire Detection Model for Underground Utility Facilities Using Deep Learning : Training Data Supplement and Bias Optimization

딥러닝 기반 지하공동구 화재 탐지 모델 개발 : 학습데이터 보강 및 편향 최적화

  • Kim, Jeongsoo (Korea BIM Research Center, Korea Institute of Civil Engineering and Building Technologies) ;
  • Lee, Chan-Woo (SPIN A WEB) ;
  • Park, Seung-Hwa (Korea BIM Research Center, Korea Institute of Civil Engineering and Building Technologies) ;
  • Lee, Jong-Hyun (SPIN A WEB) ;
  • Hong, Chang-Hee (Korea BIM Research Center, Korea Institute of Civil Engineering and Building Technologies)
  • 김정수 (한국건설기술연구원 국가BIM연구센터) ;
  • 이찬우 (스핀어웹) ;
  • 박승화 (한국건설기술연구원 국가BIM연구센터) ;
  • 이종현 (스핀어웹) ;
  • 홍창희 (한국건설기술연구원 국가BIM연구센터)
  • Received : 2020.11.02
  • Accepted : 2020.12.04
  • Published : 2020.12.31

Abstract

Fire is difficult to achieve good performance in image detection using deep learning because of its high irregularity. In particular, there is little data on fire detection in underground utility facilities, which have poor light conditions and many objects similar to fire. These make fire detection challenging and cause low performance of deep learning models. Therefore, this study proposed a fire detection model using deep learning and estimated the performance of the model. The proposed model was designed using a combination of a basic convolutional neural network, Inception block of GoogleNet, and Skip connection of ResNet to optimize the deep learning model for fire detection under underground utility facilities. In addition, a training technique for the model was proposed. To examine the effectiveness of the method, the trained model was applied to fire images, which included fire and non-fire (which can be misunderstood as a fire) objects under the underground facilities or similar conditions, and results were analyzed. Metrics, such as precision and recall from deep learning models of other studies, were compared with those of the proposed model to estimate the model performance qualitatively. The results showed that the proposed model has high precision and recall for fire detection under low light intensity and both low erroneous and missing detection capabilities for things similar to fire.

화재는 높은 비정형성으로 인해 딥러닝 모델을 이용한 영상인식 분야에서도 좋은 성능을 내기가 어려운 대상 중 하나이다. 특히 지하공동구 내 화재는 딥러닝 모델의 학습을 위한 화재 데이터 확보가 어렵고 열약한 영상 조건 및 화재로 오인할 수 있는 객체가 많아 화재 검출이 어렵고 성능이 낮다. 이러한 이유로 본 연구는 딥러닝 기반의 지하공동구 내 화재 탐지 모델을 제안하고, 제안된 모델의 성능을 평가하였다. 기존 합성곱 인공신경망에 GoogleNet의 Inception block과 ResNet의 skip connection을 조합하여 어두운 환경에서 발생되는 화재 탐지를 위한 모델 구조를 제안하였으며, 제안된 모델을 효과적으로 학습시키기 위한 방법도 함께 제시하였다. 제안된 방법의 효과를 평가하기 위해 학습 후 모델을 지하공동구 및 유사환경 조건의 화재 문제와 화재로 오인할 수 있는 객체를 포함한 이미지에 적용해 결과를 분석하였다. 또한 기존 딥러닝 기반 화재 탐지 모델의 정밀도, 검출률 지표와 비교함으로써 모델의 화재 탐지 성능을 정량적으로 평가하였다. 제안된 모델의 결과는 어두운 환경에서 발생되는 화재 문제에 대해 높은 정밀도와 검출률을 나타내었으며, 유사 화재 객체에 대해 낮은 오탐 및 미탐 성능을 가지고 있음을 보여주었다.

Keywords

References

  1. J. W. Shin, "Introduction of Recent Deep Learning Algorithms for Image Identification," The Journal of the Korea Institute of Communication Sciences, vol.34, no.7, pp.25-30, 2017. (In Korean)
  2. S. Shim, S. I. Choi, "Development on Identification Algorithm of Risk Situation around Construction Vehicle Using YOLO-v3," Journal of the Korea Academia-Industrial Cooperation Society, vol.20, no.7, pp.622-629, July 2019. (In Korean) DOI: https://doi.org/10.5762/KAIS.2019.20.7.622
  3. S. Chung, S. Moon, S. Chi, "Bridge Damage Factor Recognition from Inspection Reports Using Deep Learning," Journal of the Korean Society of Civil Engineers, vol.38, no.4, pp.612-625, August 2018. DOI: https://doi.org/10.12652/Ksce.2018.38.4.0621
  4. K. J. Kim, Y. S. Park, S. W. Park, "Development of Artificial Neural Network Model for Estimation of Cable Tension of Cable-Stayed Bridge", Journal of the Korea Academia-Industrial Cooperation Society, vol.21, no.3, pp.414-419, March 2020. (In Korean) DOI: https://doi.org/10.5762/KAIS.2020.21.3.414
  5. J. Choi, "Development of Estimation Model for Hysteresis of Friction Using Artificial Intelligent," Journal of the Korea Academia-Industrial Cooperation Society, vol.12, no.7, pp.2913-2918, July 2011. (In Korean) DOI: https://doi.org/10.5762/KAIS.2011.12.7.2913
  6. B. Lee, D. Han, "Flame and Smoke Detection Method for Early and Real-time Detection of Tunnel Fire," The Institute of Electronics Engineers of Korea-Signal Processing, vol.45, no.4, pp.59-70, July 2008. (In Korean)
  7. J. Yim, H. Park, W. Lee, M. S. Kim, Y. T. Lee, "Deep Learning Based CCTV Fire Detection System," Proceeding of the Korea Institute of Broadcast and Media Engineers, pp.139-141, 2017. (In Korean)
  8. H. S. Shin, K. B. Lee, M. J. Yim, D. G. Kim, "Development of a Deep-Learning Based Tunnel Incident Detection System on CCTVs," Journal of Korean Tunnelling and Underground Space Association, vol.19, no.6, pp.915-936, 2017. (In Korean) DOI: https://doi.org/10.9711/KTAJ.2017.19.6.915
  9. H. S. Shin, D. G. Kim, M. J. Yim, K. B. Lee, , Y. S. Oh, "A Preliminary Study for Development of an Automatic Incident Detection System on CCTV in Tunnels Based on a Machine Learning Algorithm," Journal of Korean Tunnelling and Underground Space Association, vol.19, no.1, pp.95-107, 2017. (In Korean) DOI: https://doi.org/10.9711/KTAJ.2017.19.1.095
  10. K. B. Lee, H. S. Shin, "Effect on Self-Enhancement of Deep-Learning Inference by Repeated Training of False Detection Cases in Tunnel Accident Image Detection," Journal of Korean Tunnelling and Underground Space Association, vol.21, no.3, pp.419-432, 2019. (In Korean) DOI: https://doi.org/10.9711/KTAJ.2019.21.3.419
  11. T. H. Kim (2017) Deep Learning from Scratch, p.312, Hanbit Media, pp.268-272, 2017. (In Korean)
  12. C. Szegedy, W. Lui, Y. Jia, P. Sermanet, S. Reed, D. Anguelov, D. Ertan, V. Vanhoucke, A. Rabinovich, "Going Deeper with Convolutions," In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.1-9, 2015.
  13. K. He, X. Zhang, S. Ren, J. Sun, "Deep Residual Learning for Image Recognition," In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.770-778, 2016.
  14. R. Roy. Using YOLOv3 for Real-Time Detection of PPE and Fire [Internet], Towards Data Science, c2020 [cited 2020 May 12], Available From: https://towardsdatascience.com/using-yolov3-for-real-time-detection-of-ppe-and-fire-1c671fcc0f0e (Accessed Oct. 29, 2020)
  15. W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, C.Y. Fu, A.C. Berg, "SSD: Single shot multibox detector," In Proceedings of European Conference on Computer Vision(ECCV), Amsterdam, Netherlands, pp.1-17, Oct. 2016. quoted in Park and Ko (2020) [18]
  16. S. Wu, L. Zhang, "Using popular object detection methods for real time forest fire detection," In Proceedings of the 11th International Symposium on Computational Intelligence and Design (SCID), Hangzhou, China, pp.280-284, Dec. 2018. quoted in Park and Ko (2020) [18] DOI: https://doi.org/10.1109/ISCID.2018.00070
  17. J. Redmon, A. Farhadi, "YOLOv3: Incremental Improvement," ArXiv Preprint arXiv:1804.02767, 2018. quoted in Park and Ko (2020) [18]
  18. M. Park, B.C. Ko, "Two-Step Real-Time Night-Time Fire Detection in an Urban Environment Using Static ELASTIC-YOLOv3 and Temporal Fire-Tube," Sensors, vol.20, no.8, 2202, April 2020. DOI: https://doi.org/10.3390/s20082202
  19. K. Muhammad, J. Ahmad, I. Mehmood, S. Rho, S. W. Baik, "Convolutional Neural Networks Based Fire Detection in Surveillance Videos," IEEE Transactions on Systems, Mand, and Cybernetics: Systems, vol.49, no. 7, pp.1419-1434, April 2018. DOI: https://doi.org/10.1109/TSMC.2018.2830099
  20. K. Muhammad, J. Ahmad, I. Mehmood, S. W. Baik, "Early Fire Detection Using Convolutional Neural Networks during Surveillance for Effective Disaster Management", Neurocomputing, vol.288, pp.30-42, May 2018. quoted in Muhammad et al. (2018) [19] DOI: https://doi.org/10.1016/j.neucom.2017.04.083
  21. D. Y. T. Chino, L. P. S. Avalhais, J. F. Rodrigues, A. J. M. Traina, "BoWFire: Detection of Fire in Still Images by Integrating Pixel Color and Texture Analysis," in Proc. 28th SIBGRAPI Conf. Graph., Patterns Images, pp.95-102, August 2015. quoted in Muhammad et al. (2018) [19]
  22. S. Rudz, K. Chetehouna, A. Hafiane, H. Laurent, O. Sero-Guillaume, "Investigation of a Novel Image Segmentation Method Dedicated to Forest Fire Applications," Meas. Sci. Technol., vol.24, no.7, p.075403, 2013. quoted in Muhammad et al. (2018) [19] https://doi.org/10.1088/0957-0233/24/7/075403
  23. L. Rossi, M. Akhloufi, Y. Tison, "On the Use of Stereovision to Develop a Novel Instrumentation System to Extract Geometric Fire Fronts Characteristics," Fire Safety Journal, vol.46, pp.9-20, 2011. quoted in Muhammad et al. (2018) [19] DOI: https://doi.org/10.1016/j.firesaf.2010.03.001
  24. T. Celik, H. Demirel, "Fire Detection in Video Sequences Using a Generic Color Model," Fire Safety Journal, vol.46, no.2, pp.147-158, 2009. quoted in Muhammad et al. (2018) [19] DOI: https://doi.org/10.1016/j.firesaf.2008.05.005
  25. S. Verstockt, T. Beji, P. D. Potter, S. V. Hoecke, B. Sette, B. Merci, R.V. Walle, "Video Driven Fire Spread Forecasting (f) Using Multi-Modal LWIR and Visual Flame and Smoke Data," Pattern Recognition Letters, vol.34, no.1, pp.62-69, 2013. quoted in Muhammad et al. (2018) [19] DOI: https://doi.org/10.1016/j.patrec.2012.07.018
  26. B.C. Ko, S. J. Ham, J. Y. Nam, "Modeling and Formalization of Fuzzy Finite Automata for Detection of Irregular Fire Flames," IEEE Trans. Circuits Syst. Video Technol., vol.21, no.12, pp.1903-1912, Dec. 2011. quoted in Muhammad et al. (2018) [19] DOI: https://doi.org/10.1109/TCSVT.2011.2157190