Journal of the Korea Institute of Information and Communication Engineering (한국정보통신학회논문지)

The Korea Institute of Information and Commucation Engineering (한국정보통신학회)
권호 표지
DOI QR코드

DOI QR Code

The System of Arresting Wanted Vehicles for Violent Crimes for Public Safety

국민안전을 위한 강력범죄 수배차량 검거시스템

  • Ji, Moon-Se (Department of Software Security, Graduate School of Computer & Information Technology, Korea University) ;
  • Ki, Heajeong (Research Institute, Research Institute, Bluecoms Co. Ltd.) ;
  • Ki, Chang-Min (Management Supports Department, Bluecoms Co. Ltd.) ;
  • Moon, Beom-Seob (Research Institute, Bluecoms Co. Ltd.) ;
  • Park, Sung-Geon (Research Institute, Bluecoms Co. Ltd.)
  • Received : 2021.10.18
  • Accepted : 2021.11.10
  • Published : 2021.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

The final goal of this study is to develop a system that can analyze whether a wanted vehicle is a criminal vehicle from images collected from black boxes, smartphones, CCTVs, and so on. Data collection was collected using a self-developed black box. The used data in this study has used a total of 83,753 cases such as the eight vehicle types(truck, RV, passenger car, van, SUV, bus, sports car, electric vehicle) and 434 vehicle models. As a result of vehicle recognition using YOLO v5, mAP was found to be 80%. As a result of identifying the vehicle model with ReXNet using the self-developed black box, the accuracy was found to be 99%. The result was verified by surveying field police officers. These results suggest that improving the accuracy of data labeling helps to improve vehicle recognition performance.

본 연구의 최종 목표는 블랙박스, 스마트폰, CCTV 등으로부터 수집된 영상에서 수배차량이 범죄차량인지 여부를 분석할 수 있는 시스템을 개발하는 것이다. 데이터 수집은 자체 개발된 블랙박스를 이용하였다. 본 연구에 활용된 데이터는 차량 유형 8개(트럭, RV, 승용차, 승합차, SUV, 버스, 스포츠카, 전기차)와 차량 모델 434개 등 총 83,753건의 데이터를 사용하였다. YOLO v5를 이용한 차량인식 결과, mAP가 80%로 나타났다. 자체 개발한 블랙박스를 이용하여 ReXNet으로 차량 모델을 식별한 결과, 정확도는 99%로 나타났다. 이러한 결과는 데이터 라벨링의 정확도를 개선하는 것이 차량인식 성능 향상에 도움이 된다는 것을 의미한다.

Keywords

Acknowledgement

This research is supported by "Rediscovery of the Past R&D Result" through the Ministry of Trade, Industry and Energy(MOTIE) and the Korea Institute for Advancement of Technology(KIAT) (Grant No.: (MOTIE) (P0013959, 2020))

References

  1. Y. S. Shin, S. H. Han, and J. Y. Lee, "A Study on Methods for the Domestic Diffusion of Intelligent Security Project : With a Focus on the Case of Smart City Integrated Platform," Journal of the Korea Academia-Industrial cooperation Society, vol. 20, no. 7, pp. 474-484, July. 2019. https://doi.org/10.5762/KAIS.2019.20.7.474
  2. Ministry of Land, Infrastructure and Transport agreed with the Ministry of Justice to share CCTV video information in real-time in case of an emergency [Internet]. Available: http://www.molit.go.kr/mtc/USR/N0201/m_36770/dtl.jsp?lcmspage=50&id=95081884.
  3. C. J. Sim, S. H. Yoo, and W. I. Kim, "File Database and Search Algorithm for Efficient Search of Car Number," KIPS transactions on software and data engineering, vol. 8, no. 10, pp. 391-396, Aug. 2019.
  4. B. H. Kim, Y. J. Han, and H. S. Han, "Robust Scheme of Segmenting Characters of License Plate on Irregular Illumination Condition," Journal of the Korea Society of Computer and Information, vol. 14, no. 11, pp. 61-71, Dec. 2009.
  5. S. J. Dang and E. T. Kim, "Robust Motorbike License Plate Detection and Recognition using Image Warping based on YOLOv2," Journal of broadcast engineering, vol. 24, no. 5, pp. 713-725, Sep. 2019. https://doi.org/10.5909/JBE.2019.24.5.713
  6. E. D. Cubuk, B. Zoph, D. Mane, V. Vasudevan, and Q. V. Le, "Autoaugment: Learning augmentation strategies from data," in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 113-123, Apr. 2019.
  7. S. W. Lim and G. M. Park, "Development of Python-based Annotation Tool Program for Constructing Object Recognition Deep-Learning Model," Journal of broadcast engineering, vol. 25, no. 3, May. 2020.
  8. K. Simonyan and A. Zisserman, "Very deep convolutional networks for large-scale image recognition," arXiv preprintarXiv:1409, pp. 1556, 2014.
  9. G. Huang, Z. Liu, L. Maaten, and K. Q. Weinberger, "Densely Connected Convolutional Networks," in Proceeding of the 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops, Honolulu, HI, USA, pp. 2261-2269, 2017.
  10. W. B, Shin, W. H. Lee, and C. H. Hwang, "A Study on the Determinants of crime arrest rate," Korean Police Studies Review, vol. 13, no. 4, pp. 89-114, Dec. 2014.
  11. D. W. Min, H. S. Lim, and J. H. Gwak, "Improved method of license plate detection and recognition facilitated by fast super-resolution GAN," Smart Media Journal, vol. 9, no. 4, pp. 134-143, Dec. 2020.
  12. J. Redmon and A. Farhadi, "YOLO9000: Better, Faster, Stronger," IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 6517-6525, June. 2017.
  13. H. Li, P. Wang, and C. Shen, "Towards end-to-end car license plates detection and recognition with deep neural networks," arxiv, Sep. 2017.
  14. D. H. Park and H. I. Kim, "Improved Object Recognition using Multi-view Camera for ADAS," Smart Media Journal, vol. 24, no. 4, pp. 573-579, Jul. 2019.
  15. J. Redmon and A. Farhadi, "YOLOv3: An Incremental Improvement," Computing Research Repository(CoRR), 2018.
  16. H. J. Lee, E. J. Lee, S. H. Park, S. Y. Ihm, and Y. H. Park, "YOLOv3-based Vehicle Detection and Counting Method through Video," in Proceeding of the Annual Conference of Korea International Processing Society, Jeju-do, pp. 935-938, 2019.
  17. L. He, X. Liao, W. Liu, X. Liu, P. Cheng, and T. Mei, "FastReID: A Pytorch Toolbox for General Instance Re-identification," JD AI Research, pp. 2-7, Jun. 2020.
  18. S. J. Leem, T. J. Kim, C. H. Lee, and S. B. Yoo, "Image Super-Resolution for Improving Object Recognition Accuracy," Journal of the Korea Institute of Information and Communication Engineering, vol. 25, no. 6, pp. 774-784, Jun. 2021. https://doi.org/10.6109/JKIICE.2021.25.6.774
  19. K. C. Lee, G. Wang, and S. Y. Shin, "Structure, Method, and Improved Performance Evaluation Function of SRCNN and VDSR," Journal of the Korea Institute of Information and Communication Engineering, vol. 25, no. 4, pp. 543-548, Apr. 2021. https://doi.org/10.6109/JKIICE.2021.25.4.543
  20. W. T. Freeman, T. T. Jonesm, and E. C. Pasztor, "Example-based Super-resolution," Journal of IEEE Transaction on Computer Graphics and Application, vol. 22, no. 2, pp. 56-65, Mar. 2002.
  21. H. J. Lee, H. K. Shin, G. S. Choi, and S. I. Jin, "Performance Analysis of Deep Learning-based Image Super Resolution Methods," Journal of embedded systems and applications, vol. 15, no. 2, pp. 61-70, Apr. 2020.
  22. Compound-scaled object detection models trained on the COCO dataset [Internet]. Available: https://pytorch.org/hub/ultralytics_yolov5/.
  23. The AI model proposed a model design rule to prevent the problem of Representational Bottleneck that may occur on CNN [Internet]. Available: https://github.com/clovaai/rexnet.