Convergence Security Journal (융합보안논문지)

Korea convergence Security Association (한국융합보안학회)
권호 표지
DOI QR코드

DOI QR Code

Development of segmentation-based electric scooter parking/non-parking zone classification technology

Segmentation 기반 전동킥보드 주차/비주차 구역 분류 기술의 개발

  • 조용현 (부산대학교 정보컴퓨터공학부) ;
  • 최진영 (동신대학교 기계공학과)
  • Received : 2023.11.21
  • Accepted : 2023.12.26
  • Published : 2023.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

This paper proposes an AI model that determines parking and non-parking zones based on return authentication photos to address parking issues that may arise in shared electric scooter systems. In this study, we used a pre-trained Segformer_b0 model on ADE20K and fine-tuned it on tactile blocks and electric scooters to extract segmentation maps of objects related to parking and non-parking areas. We also presented a method to perform binary classification of parking and non-parking zones using the Swin model. Finally, after labeling a total of 1,689 images and fine-tuning the SegFomer model, it achieved an mAP of 81.26%, recognizing electric scooters and tactile blocks. The classification model, trained on a total of 2,817 images, achieved an accuracy of 92.11% and an F1-Score of 91.50% for classifying parking and non-parking areas.

본 논문은 공유형 전동킥보드 시스템 운영 시, 관리 상 발생할 수 있는 주차 문제를 해결하기 위해 반납 인증사진으로 주차, 비주차 구역을 판단하는 AI모델을 제시한다. 본 연구에서는 주차/비주차 구역 배경 관련 객체를 판별하기 위해 ADE20K에 Pre_trained된 Segfomer_b0 모델과 점자블록, 전동킥보드에 Fine_tuning한 Segfomer_b0 모델을 통해 주차/비주차에 관련된 객체의 Segmentation map을 추출하고, Swin 모델을 통해 주차/비주차 구역을 이진 분류하는 방법을 제시하였다. 최종적으로 총 1,689장을 직접 라벨링한 후 진행한 Fine_tuning SegFomer 모델은 mAP가 81.26% 수준으로 전동킥보드와 점자블록을 인식하였으며, 총 2,817장을 훈련한 Classification 모델은 92.11%의 정확도와 91.50%의 F1-Score로 주차구역과 비주차 구역을 분류하는 것이 가능하였다.

Keywords

Acknowledgement

본 논문은 2023년도 교육부의 재원으로 한국연구재단의 지원을 받아 수행된 지자체-대학 협력기반 지역혁신 사업의 결과입니다.(2021RIS-002)

References

  1. The Korea Transport Institute, "A Study Improving Leislation for Promoting Utilization of Personal Mobility and its Safety Management", Research report, Feb. 2020.
  2. The Korea Research Institute for Human Settlements, "U.S. shared electric sccoter management policy and implications", Research report, Nov. 2019.
  3. "서울특별시 정차.주차위반차량 견인 등에 관한 조례 일부개정조례안", Oct 2020.
  4. "전동킥보드 공유서비스 안전실태조사", Aug 2021.
  5. Ashish. Vaswani, Noam. Shazzer, Niki. Parmar, Jakob. Uszkoreit, Llion. Jones, Aidan N. Gomez, Lukasz. Kaiser and Lllia. Polosukhin, "Attention Is All You Need" axXiv preprint arXiv:1706.03762v5, Dec. 2027.
  6. Alexey. Dosovitkiy, Lucas. Beyer, Alexander. Kolesnikov, Dirk. Weissenborn, Xiaohua. Zhai, Thomas. Unterthiner, Mostafa. Dehghani, Matthias. Minderer, Georg. Heigold, Sylvain. Gelly, Jakob. Uszkoreit, and Neil. Houlsby, "AN IMAGE IS WORTH 16X16 WORDS : TRANSFORMERS FOR IMAGE RECOGNITION AT SCALE" arXiv:2010.11929v2, Jun 2021.
  7. Enze. Xie, Wenhai. Wang, Zhiding. Yu, Anima. Anandkumar, Jose M. Alvarez, and Ping. Luo, "SegFormer : Simple and Efficient Design for Semantic Segmentation with Transformers" arXiv:2015.15203v3, Oct 2021.
  8. Ze. Liu, Yutong. Lin, Yue. Cao, Han. Hu, Yixuan. Wei, Zheng. Zhang, Stephen. Lin, and Baining. Gue, "Swin Transformer : Hierarchical Vision Transformer using Shifted Windows", arXiv preprint arXiv:2103.14030v2, Aug 2021.
  9. Rudra. PK. Poudel, Stephan. Liwicki, and Roberto. Cipolla, "Fast-SCNN:Fast Semantic Segmentation Network", arXivn:1902.04502v1, Feb 2019.
  10. Chao, Ping, Chao-Yang Kao, Yu-Shan Rua n, Chien-Hsiang Huang, and Youn-Long Lin. "Hardnet:A low memory traffic network" In Proceedings of the IEEE International Conference on Computer Vision, pp. 3552-3561. Sep 2019.
  11. Fan, Mingyuan, Shenqi Lai, Junshi Huang, Xiaoming Wei, Zhenhua Chai, Junfeng Luo, and Xiaolin Wei, "Rethinking BiSeNet For Real-time Semantic Segmentation." In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 9816-9725. Apr 2021.
  12. Ronneberger O, Fischer P, Brox T, "U-net: Convolutional networks for biomedical image segmentation" International Conference on Medical image computing and computer-assisted intervention, Springer, Cham, 2015:234-241. May 2015.