Journal of Korea Society of Industrial Information Systems (한국산업정보학회논문지)

Korea Society of Industrial Information Systems (한국산업정보학회)
권호 표지
DOI QR코드

DOI QR Code

Computer Vision-Based Car Accident Detection using YOLOv8

YOLO v8을 활용한 컴퓨터 비전 기반 교통사고 탐지

  • Received : 2023.10.27
  • Accepted : 2023.11.25
  • Published : 2024.02.29
Download PDF
⟨ Previous Next ⟩

Abstract

Car accidents occur as a result of collisions between vehicles, leading to both vehicle damage and personal and material losses. This study developed a vehicle accident detection model based on 2,550 image frames extracted from car accident videos uploaded to YouTube, captured by CCTV. To preprocess the data, bounding boxes were annotated using roboflow.com, and the dataset was augmented by flipping images at various angles. The You Only Look Once version 8 (YOLOv8) model was employed for training, achieving an average accuracy of 0.954 in accident detection. The proposed model holds practical significance by facilitating prompt alarm transmission in emergency situations. Furthermore, it contributes to the research on developing an effective and efficient mechanism for vehicle accident detection, which can be utilized on devices like smartphones. Future research aims to refine the detection capabilities by integrating additional data including sound.

자동차 사고는 차량 간의 충돌로 인해 발생되며, 이로 인해 차량의 손상과 함께 인적, 물적 피해가 유발된다. 본 연구는 CCTV에 의해 촬영되어 YouTube에 업로드된 차량사고 동영상으로 부터 추출된 2,550개의 이미지 프레임을 기반으로 차량사고 탐지모델을 개발하였다. 전처리를 위해 roboflow.com을 사용하여 바운딩 박스를 표시하고 이미지를 다양한 각도로 뒤집어 데이터 세트를 증강하였다. 훈련에서는 You Only Look Once 버전 8 (YOLOv8) 모델을 사용하였고, 사고 탐지에 있어서 평균 0.954의 정확도를 달성하였다. 제안된 모델은 비상시에 경보 전송을 용이하게 하는 실용적 의의를 가지고 있다. 또한, 효과적이고 효율적인 차량사고 탐지 메커니즘 개발에 대한 연구에 기여하고 스마트폰과 같은 기기에서 활용될 수 있다. 향후의 연구에서는 소리와 같은 추가 데이터의 통합을 포함하여 탐지기능을 정교화하고자 한다.

Keywords

Acknowledgement

This study was supported by a grant from Korea Tech. & Info. Promotion Agency (No. S3302201).

References

  1. Abou, L., Fliflet, A., Hawari, L., Presti, P., Sosnoff, J. J., Mahajan, H. P., and Rice, L. A. (2022). Sensitivity of Apple Watch fall detection feature among wheelchair users, Assistive technology, 34(5), 619-625.  https://doi.org/10.1080/10400435.2021.1923087
  2. CDC. (2019). Road Traffic Injuries and Deaths: A Global Problem tical Programming Tools, https://www.cdc.gov/injury/features/global-road-safety/index.html 
  3. Choi, D. H., Lee, J., & Lee, D. (2022). A study on the detection of pedestrians in crosswalks using multi-spectrum. Journal of the Korea Industrial Information Systems Research, 27(1), 11-18. 
  4. Choi, S., Lee, K., Kim, K., & Kwak, S. (2019). Lane departure warning system using deep learning. Journal of the Korea Industrial Information Systems Research, 24(2), 25-31. 
  5. Desai, R., Jadhav, A., Sawant, S., and Thakur, N. (2021). Accident Detection Using ML and AI Techniques, Engpaper Journal.
  6. Gandhi R. (2018). R-CNN, Fast R-CNN, Faster R-CNN, YOLO - Object Detection Algorithms,https://towardsdatascience.com/rcnn-fast-r-cnn-faster-r-cnn-yolo-object-detection-algorithms-36d53571365e 
  7. Girshick, R. (2015). Fast r-cnn. In proceedings of the IEEE international conference on computer vision (pp. 1440-1448). 
  8. Gour, D., and Kanskar, A. (2019). Optimised YOLO: algorithm for CPU to detect road traffic accident and alert system, International. J . Eng. Res. Technol, 8. 160-163. 
  9. Gupta, U., Kim, Y. G., Lee, S., Tse, J., Lee, H. H. S., Wei, G. Y., and Wu, C. J. (2021, February). Chasing carbon: The elusive environmental footprint of computing. In 2021 IEEE International Symposium on High-Performance Computer Architecture (HPCA) (pp. 854-867). IEEE. 
  10. Jacob S, Francesco. (2020). What is Mean Average Precision (mAP) in Object Detection?, https://blog.roboflow.com/mean-average-precision/ 
  11. Jiang, Peiyuan, Daji Ergu, Fangyao Liu, Ying Cai, and Bo Ma. (2022). A Review of Yolo algorithm developments, P rocedia Computer Science, 199. 1066-1073. 
  12. Ki, Y. K., and Lee, D. Y. (2007). A traffic accident recording and reporting model at intersections, IEEE Transactions on Intelligent Transportation Systems, 8(2), 188-194.  https://doi.org/10.1109/TITS.2006.890070
  13. Lee, M. H., Nam, K. W., & Lee, C. W. (2019). Crack Detection on the Road in Aerial Image using Mask R-CNN. Journal of the Korea Industrial Information Systems Research, 24(3), 23-29. 
  14. Liu, B., He, F., Du, S., Li, J., & Liu, W. (2022). An advanced YOLOv3 method for small object detection. arXiv preprint arXiv:2212.02809.. 
  15. Nasr, E., Kfoury, E., and Khoury, D. (2016). An IoT approach to vehicle accident detection, reporting, and navigation, In 2016 IEEE international multidisciplinary conference on engineering technology, 231-236. 
  16. Redmon, J., & Farhadi, A. (2018). Yolov3: An incremental improvement. arXiv preprint arXiv:1804.02767. 
  17. Redmon, J., Divvala, S., Girshick, R., and Farhadi, A. (2016). You only look once: Unified, real-time object detection, In Proceedings of the IEEE conference on computer vision and pattern recognition, 779-788. 
  18. Ren, S., He, K., Girshick, R., & Sun, J. (2015). Faster r-cnn: Towards real-time object detection with region proposal networks. Advances in neural information processing systems, 28. 
  19. Sharma, S., and Sebastian, S. (2019). IoT based car accident detection and notification algorithm for general road accidents, International Journal of Electrical & Computer Engineering, 9(5). 2088-8708. 
  20. Terven, J., and Cordova-Esparza, D (2023). A comprehensive review of YOLO: From YOLOv1 to YOLOv8 and beyond, arXiv preprint, arXiv:2304.00501. 
  21. Tian, D., Zhang, C., Duan, X., and Wang, X. (2019). An automatic car accident detection method based on cooperative vehicle infrastructure systems, IEEE Access, 7. 127453-127463.  https://doi.org/10.1109/ACCESS.2019.2939532
  22. Topinkatti, A., Yadav, D., Kushwaha, V. S., and Kumari, A. (2015). Car accident detection system using GPS and GSM, International Journal of Engineering Research and General Science, 3(3). 1025-1033. 
  23. Ultralytics. (2023). Ultralytics YOLOv8 documentation, https://docs.ultralytics.com