전기전자학회논문지 (Journal of IKEEE)

  • 제25권4호
  • /
  • Pages.750-755
  • /
  • 2021
  • /
  • 1226-7244(pISSN)
  • /
  • 2288-243X(eISSN)
한국전기전자학회 (Institute of Korean Electrical and Electronics Engineers)
권호 표지
DOI QR코드

DOI QR Code

YOLOv4를 이용한 차량파손 검출 모델 개선

Improving the Vehicle Damage Detection Model using YOLOv4

  • Jeon, Jong Won (Dept. of Information and Communications Engineering, Hankuk University of Foreign Studies) ;
  • Lee, Hyo Seop (Dept. of Information and Communications Engineering, Hankuk University of Foreign Studies) ;
  • Hahn, Hee Il (Dept. of Information and Communications Engineering, Hankuk University of Foreign Studies)
  • 투고 : 2021.11.30
  • 심사 : 2021.12.31
  • 발행 : 2021.12.31
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

본 논문에서는 YOLOv4를 이용하여 차량의 부위별 파손현황을 검출하는 기법을 제안한다. 제안 알고리즘은 YOLOv4를 통해 차량의 부위와 파손을 각각 학습시킨 후 검출되는 바운딩 박스의 좌표 정보들을 추출하여 파손과 차량부위의 포함관계를 판단하는 알고리즘을 적용시켜 부위별 파손현황을 도출한다. 또한 성능비교의 객관성을 위하여 동일분야의 VGGNet을 이용한 기법, 이미지 분할과 U-Net 모델을 이용한 기법, Weproove.AI 딥러닝 모델 등을 대조 모델로 포함한다. 이를 통하여 제안 알고리즘의 성능을 비교, 평가하고 검출 모델의 개선 방안을 제안한다.

This paper proposes techniques for detecting the damage status of each part of a vehicle using YOLOv4. The proposed algorithm learns the parts and their damages of the vehicle through YOLOv4, extracts the coordinate information of the detected bounding boxes, and applies the algorithm to determine the relationship between the damage and the vehicle part to derive the damage status for each part. In addition, the technique using VGGNet, the technique using image segmentation and U-Net model, and Weproove.AI deep learning model, etc. are included for objectivity of performance comparison. Through this, the performance of the proposed algorithm is compared and evaluated, and a method to improve the detection model is proposed.

키워드

과제정보

This work was supported by Hankuk University of Foreign Studies Research Fund.

참고문헌

  1. K. Patil, M. Kulkarni, A. Sriraman, and S. Karande, "Deep learning based car damage classification," 16th IEEE International Conference on Machine Learning and Applications, 2017. DOI: 10.1109/ICMLA.2017.0-179
  2. H. S. Malik, M. Dwivedi, S. N. Omakar, S. R. Samal, A. Rathi, E. B. Monis, B. Khanna, and A. Tiwari, "Deep learning based car damage classification and detection," EasyChair preprints, 2020. DOI: 10.1109/ICMLA.2017.0-179
  3. P. Rakshata, H. V. Padma, M. Pooja, H. V. Yashaswini, and V. Karthik, "Car damage detection and analysis using deep learning algorithm for automotive," International Journal of Scientific Research & Engineering Trends, Vol.5, Issue.6, 2019.
  4. P. M. Kyu and K. Woraratpanya, "Car damage assessment based on VGG models," JSCI8, 2021.
  5. WeProov.AI, "AI for vehicle inspection, damages & claims management," 2020.
  6. SOCAR Tech, "Developing car damage detection model using semantic segmentation," https://tech.socarcorp.kr/data/2020/02/13/car-damage-segmentation-model.html, Feb. 2020.
  7. K. Simonyan and A. Zisserman, "Very Deep Convolutional Networks for Large-Scale Image Recognition," arXiv:1409.1556 [cs.CV], Sep. 2014.
  8. OpenGenus IQ, "Understanding the VGG19 A rchitecture," https://iq.opengenus.org/vgg19-architecture/, 2018.
  9. O. Ronneberger, P. Fischer, and T. Brox, "U-Net: Convolutional Networks for Biomedical Image Segmentation," arXiv:1505.04597 [cs.CV], 2015.
  10. A. Bochkovskiy, C. Y. Wang, and H. M. Liao, "YOLOv4: Optimal speed and accuracy of object detection," arXiv:2004.10934v1, 2020.
  11. Car damage dataset, https://github.com/jennyluciav/car-damage-dataset.