한국전자통신학회논문지 (The Journal of the Korea institute of electronic communication sciences)

한국전자통신학회 (Korea Institute of Electronic Communication Science)
권호 표지
DOI QR코드

DOI QR Code

딥러닝 알고리즘 기반 교통법규 위반 공익신고 영상 분석 시스템

Analysis System for Public Interest Report Video of Traffic Law Violation based on Deep Learning Algorithms

  • 투고 : 2022.11.14
  • 심사 : 2023.02.17
  • 발행 : 2023.02.28
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

고화질 블랙박스의 확산과 '스마트 국민제보', '안전신문고' 등 모바일 애플리케이션의 도입에 따른 영향으로 교통법규 위반 공익신고가 급증하였으며, 이로 인해 이를 처리할 담당 경찰 인력은 부족한 상황이 되었다. 본 논문에서는 교통법규 위반 공익신고 영상 중, 가장 많은 비중을 차지하는 차선위반에 대해 딥러닝 알고리즘을 활용하여 자동 검출할 수 있는 시스템의 개발내용에 관해 기술한다. 본 연구에서는 YOLO 모델과 Lanenet 모델을 사용하여 차량과 실선 객체를 인식하고 deep sort 알고리즘을 사용하여 객체를 개별로 추적하는 방법, 그리고 차량 객체의 바운딩 박스와 실선 객체의 범위가 겹치는 부분을 인식하여 진로변경 위반을 검출하는 방법을 제안한다. 본 시스템을 통해 신고된 영상에 대해 교통법규 위반 여부를 자동 분석해줌으로써 담당 경찰 인력 부족난을 해소할 수 있을 것으로 기대한다.

Due to the spread of high-definition black boxes and the introduction of mobile applications such as 'Smart Citizens Report' and 'Safety Report', the number of public interest reports for violations of Traffic Law has increased rapidly, resulting in shortage of police personnel to handle them. In this paper, we describe the development of a system that can automatically detect lane violations which account for the largest proportion of public interest reporting videos for violations of traffic laws, using deep learning algorithms. In this study, a method for recognizing a vehicle and a solid line object using a YOLO model and a Lanenet model, a method for tracking an object individually using a deep sort algorithm, and a method for detecting lane change violations by recognizing the overlapping range of a vehicle object's bounding box and a solid line object are described. Using this system, it is expected that the shortage of police personnel in charge will be resolved.

키워드

과제정보

이 논문은 2022년도 동서대학교 "Dongseo Cluster Project" 지원에 의하여 이루어진 것임 (DSU-20220002)

참고문헌

  1. Anti-Corruption & Civil Rights Commission, "5.41 million public interest reports were received in 2021," press release, July 2022.
  2. The National Police Agency, "Traffic Violation Report Surges," press release, July 2022.
  3. D. Kim and S. Kim, "A Study on Risk Situation Awareness Using OpenCV," J. of the Korea Institute of Electronic Communication Sciences, vol. 16, no. 2, 2021, pp. 211-218.
  4. M. Moon and G. Lee, "Development a Meal Support System for the Visually Impaired Using YOLO Algorithm," J. of the Korea Institute of Electronic Communication Sciences, vol. 16, no. 5, 2021, pp. 1001-1010. https://doi.org/10.13067/JKIECS.2021.16.5.1001
  5. M. Moon, M. Kim, and C. Han, "Expiration Date Notification System Based on YOLO and OCR algorithms for Visually Impaired Person," J. of the Korea Institute of Electronic Communication Sciences, vol. 16, no. 6, 2021, pp. 1329-1338.
  6. Ministry of Science and ICT, "National Strategy for Artificial Intelligence", press release, Dec. 2019.
  7. S. Kim, T. Kang, and J. Kim, "Automation for Progress Monitoring based by Vision Technology in a Modular Building Factory," J. of the Architectural Institute of Korea, vol. 37, no. 6, 2021, pp. 221-229.
  8. G. Jeong and Y. Choi, "An Analysis of Blur Processing Scene Image of Event/Accident Broadcast News Using Blur Processing Automation Detection Method," In Proc. Conf. Korean Association for Broadcasting and Telecommunication Studies (KABS), Jeju, South Korea, Apr. 2022, pp. 117-118.
  9. J. Lim and S. Youn, "A Study on Traffic Violation Surveillance System Based on Edge AI-Detecting a car which cuts in left-turn waiting lines," In Proc. Conf. Korean Institute of Communications and Information Sciences, Pyeongchang, South Korea, Aug. 2020, pp. 1315-1316.
  10. Y. Yao, M. Xu, Y. Wang, D. J. Crandall, and E. M. Atkins, "Unsupervised Traffic Accident Detection in First-Person Videos," In 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Macao, China, Nov. 2019, pp. 273-280.