The Journal of The Korea Institute of Intelligent Transport Systems (한국ITS학회 논문지)

The Korea Institute of Intelligent Transport Systems (한국ITS학회)
권호 표지
DOI QR코드

DOI QR Code

Development of a Emergency Situation Detection Algorithm Using a Vehicle Dash Cam

차량 단말기 기반 돌발상황 검지 알고리즘 개발

  • Sanghyun Lee (Dept. of Transportation Eng., Univ. of Ajou) ;
  • Jinyoung Kim (Dept. of D.N.A. Plus Fusion., Univ. of Ajou) ;
  • Jongmin Noh (Dept. of D.N.A. Plus Fusion., Univ. of Ajou) ;
  • Hwanpil Lee (Division of Transportation Research, Korea Expressway Corporation Research Institute) ;
  • Soomok Lee (Dept. of AI Mobility Eng., Univ. of Ajou) ;
  • Ilsoo Yun (Dept. of Transportation Eng., Univ. of Ajou)
  • 이상현 (아주대학교 교통공학과) ;
  • 김진영 (아주대학교 D.N.A.플러스융합학과) ;
  • 노종민 (아주대학교 D.N.A.플러스융합학과) ;
  • 이환필 (한국도로공사 도로교통연구원) ;
  • 이수목 (아주대학교 AI모빌리티공학과) ;
  • 윤일수 (아주대학교 교통시스템공학과)
  • Received : 2023.05.01
  • Accepted : 2023.06.30
  • Published : 2023.08.31
Download PDF
⟨ Previous Next ⟩

Abstract

Swift and appropriate responses in emergency situations like objects falling on the road can bring convenience to road users and effectively reduces secondary traffic accidents. In Korea, current intelligent transportation system (ITS)-based detection systems for emergency road situations mainly rely on loop detectors and CCTV cameras, which only capture road data within detection range of the equipment. Therefore, a new detection method is needed to identify emergency situations in spatially shaded areas that existing ITS detection systems cannot reach. In this study, we propose a ResNet-based algorithm that detects and classifies emergency situations from vehicle camera footage. We collected front-view driving videos recorded on Korean highways, labeling each video by defining the type of emergency, and training the proposed algorithm with the data.

전방 낙하물과 같은 돌발상황이 발생했을 때 신속하고 적절한 정보 제공은 도로 위 이용자들의 편의를 가져다주고 2차 교통사고 또한 효과적으로 줄일 수 있다. 도로 상의 돌발상황은 현재 국내에서 루프 검지기나 CCTV 등 ITS 기반 검지 체계를 사용하여 주로 검지하고 있다. 이러한 방식은 검지기의 검지 구간에서의 도로 위 데이터만을 얻을 수 있다. 때문에, 기존 ITS 기반 검지체계의 공간적 음영구간에서 돌발상황을 찾아내기 위하여 새로운 검지 수단이 필요하다. 이에 본 연구에서는 차량 내 설치된 단말기에서 촬영된 영상으로부터 돌발상황을 검지 및 분류하는 ResNet 기반 알고리즘을 제안한다. 국내 고속도로 전방 주행영상을 수집하였고, 돌발상황 유형을 클래스로 정의하여 각 데이터를 라벨링한 후, 제안한 알고리즘으로 데이터를 학습시켰다. 학습 결과, 개발한 알고리즘은 데이터 수가 상대적으로 적었던 일부 클래스를 제외하고 정의한 돌발상황 클래스에 대하여 높은 검지율을 보였다.

Keywords

Acknowledgement

본 논문은 한국도로공사에서 발주한 "2030 고속도로 환경을 고려한 교통정보 수집체계 진단 및 발전방안 수립 연구"의 일환으로 작성되었습니다.

References

  1. Baek, S. K.(2017), "Future Highway Prospects and Road Infrastructure Policy Direction", The Korea Spatial Planning Review, vol. 428, pp.15-19.
  2. Baek, S. K., Oh, C. S., Kang, J. G. and Nam, D. H.(2005a), "Assessments of freeway incidents management systems", Journal of the Korean Society of Civil Engineers, pp.4075-4078.
  3. Baek, S. K., Oh, C. S., Kang, J. G. and Nam, D. H.(2005b), "A guideline for freeway incident management manual", The Journal of The Korea Institute of Intelligent Transport Systems, vol. 4, no. 3, pp.61-72.
  4. Elhassouny, A. and Smarandache, F.(2019), "Trends in deep convolutional neural Networks architectures: A review", 2019 International Conference of Computer Science and Renewable Energies(ICCSRE), pp.1-8.
  5. He, K., Zhang, X., Ren, S. and Sun, J.(2016), "Deep residual learning for image recognition", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.770-778.
  6. Kang, Y. S.(2009), "ITS/Telematics", Communications of the Korean Institute of Information Scientists and Engineers, vol. 27 no. 9, pp.10-17.
  7. Ki, Y. K. and Kim, Y. H.(2020), "A travel speed prediction model for incident detection based on traffic CCTV", Journal of Industrial Convergence, vol. 18, no. 3, pp.53-61. https://doi.org/10.22678/JIC.2020.18.3.053
  8. Kim, B. J., Yim, J. H. and Kim, J. M.(2020), Highway driving dataset for semantic video segmentation, arXiv preprint arXiv:2011.00674.
  9. Kim, D. H.(2005), "Automatic Incident Detection Using Machine Learning", Seoul Studies, vol. 6, no. 1, pp.71-80. https://doi.org/10.23129/SEOULS.6.1.200503.71
  10. Lee, K. Y., Seo, I. K., Park, M. S. and Jang, M. S.(2012), "A Study on the Influencing Factors for Incident Duration Time by Expressway Accident", International Journal of Highway Engineering, vol. 14, no. 1, pp.85-94.
  11. Narayanan, A., Dwivedi, I. and Dariush, B.(2019), "Dynamic traffic scene classification with space-time coherence", 2019 International Conference on Robotics and Automation(ICRA), pp.5629-5635.
  12. Ozdamar, L., Aksu, D. T. and Ergunes, B.(2014), "Coordinating debris cleanup operations in post disaster road networks", Socio-Economic Planning Sciences, vol. 48, no. 4, pp.249-262. https://doi.org/10.1016/j.seps.2014.08.001
  13. Phung, V. H. and Rhee, E. J.(2019), "A high-accuracy model average ensemble of convolutional neural networks for classification of cloud image patches on small datasets", Applied Sciences, vol. 9, no. 21, p.4500.
  14. Prykhodchenko, R. and Skruch, P.(2022), "Road scene classification based on street-level images and spatial data", Array, vol. 15, p.100195.
  15. Sihag, G., Parida, M. and Kumar, P.(2022), "Travel Time Prediction for Traveler Information System in Heterogeneous Disordered Traffic Conditions Using GPS Trajectories", Sustainability, vol. 14, no. 16, p.10070.
  16. Society Automotive Engineers International, ITIS Phrase Lists (International Traveler Information Systems), https://www.sae.org/standards/content/j2540/2_202012, 2023.01.15.
  17. Westell, J.(2019), Multi-Task Learning using Road Surface Condition Classification and Road Scene Semantic Segmentation, Master's Thesis, Linkoping University. https://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-157403, 2023.02.21.
  18. Wu, F., Yan, S., Smith, J. S. and Zhang, B.(2021), "Deep multiple classifier fusion for traffic scene recognition", Granular Computing, vol. 6, pp.217-228. https://doi.org/10.1007/s41066-019-00182-6
  19. Yao, J., Fidler, S. and Urtasun, R.(2012), "Describing the scene as a whole: Joint object detection, scene classification and semantic segmentation", 2012 IEEE Conference on Computer Vision and Pattern Recognition, pp.702-709.
  20. Zeng, D., Liao, M., Tavakolian, M., Guo, Y., Zhou, B., Hu, D., Pietikainen, M. and Liu, L.(2021), Deep learning for scene classification: A survey, arXiv preprint arXiv:2101.10531.