Automatic Detection of Anchors of Temporary Facility Using YOLOv8n

Junwoo Shin; Taehee Lee; Wonkyun Seok; Boohyun Nam

doi:10.14481/jkges.2026.27.5.37

All Issue

2026 Vol.27, Issue 5 Preview Page

Automatic Detection of Anchors of Temporary Facility Using YOLOv8n YOLOv8n을 활용한 굴착 가시설 앵커 자동 탐지 연구

1 May 2026. pp. 37-43

PDF XML

Abstract

At civil engineering temporary construction sites, anchors are structural elements that ensure stability during ground excavation and earth-retaining work. Management standards have been established to monitor their installation status. However, as deep excavation hinder human access, there is an increasing need for systems that can automatically detect and evaluate anchors remotely. Accordingly, this study proposes a system for automatically detecting anchors at civil engineering temporary construction sites using YOLOv8n, a deep learning–based lightweight object detection model. An anchor-specific dataset was created by directly collecting and processing field image data, and the model’s performance was evaluated under various lighting and angle conditions at the site. Experimental results demonstrated that the YOLOv8n model achieved high detection accuracy of 0.995 at mAP@50 and 0.827 at mAP@50:95, along with a rapid inference speed of 2.765 ms per image. The model exhibited stable anchor detection performance even in the complex background environments typical of civil engineering temporary construction sites. This study highlights the potential of deep learning–based vision systems for safety management in the civil engineering sector and confirms their viability as foundational technology for automated field monitoring.

Keywords

Earth anchor

Temporary facility

YOLOv8n

Object detection

Deep learning

Maintenance monitoring

토목 가시설 현장에서의 앵커는 지반 굴착 및 흙막이 공사 중 구조물의 안정성을 유지하는 구조 부재로써 설치 상태를 모니터링 하도록 관리 기준이 마련되어 있다. 그러나 대심도 굴착으로 인해 사람이 접근하기 어려워짐에 따라 원격으로 자동 탐지 및 평가하는 시스템의 필요성이 증가하고 있다. 이에 본 연구는 딥러닝 기반 경량 객체 탐지 모델인 YOLOv8n을 활용하여 토목 가시설 현장에서 앵커를 자동으로 탐지하는 시스템을 제안한다. 현장 이미지 데이터를 직접 수집·가공하여 앵커 특화 데이터셋을 구축 및 학습하였으며, 현장의 다양한 조명, 각도 조건을 바탕으로 모델의 성능을 평가하였다. 실험 결과, YOLOv8n 모델은 mAP@50 기준 0.995, mAP@50:95 기준 0.827의 높은 탐지 정확도와 이미지당 2.765 ms의 지연속도로 높은 추론 속도를 기록하였으며, 토목 가시설 현장의 복잡한 배경 환경에서도 앵커를 안정적으로 탐지하는 성능을 나타냈다. 본 연구는 딥러닝 기반 비전 시스템을 토목 분야 안전 관리에 적용할 수 있는 가능성을 제시하며, 현장 모니터링 자동화의 기반 기술로 활용될 수 있음을 확인하였다.

키워드

어스 앵커

토목 가시설

YOLOv8n

객체 탐지

딥러닝

유지관리 모니터링

References

Çiftçi, M., Türkdamar, M. U. and Öztürk, C. (2024), Leveraging YOLO models for safety equipment detection on construction sites, Journal of Computing Theories and Applications, Vol. 1, No. 4, pp. 492~506.

10.62411/jcta.10453

DJI Official (2026), Camera drones product list, https://store.dji.com/kr/list/camera-drones?tab=101&page=1&category=101%3D101.201.305&from=site-nav&sort=recommendation (Accessed on April 03, 2026) (In Korean).

Duan, R., Deng, H., Tian, M., Deng, Y. and Lin, J. (2022), SODA: A large-scale open site object detection dataset for deep learning in construction, Automation in Construction, 142, 104499.

10.1016/j.autcon.2022.104499

Feng, R., Miao, Y. and Zheng, J. (2024), A YOLO-based intelligent detection algorithm for risk assess-ment of construction sites, Journal of Intelligent Construction, Vol. 2, No. 4, pp. 1~18.

10.26599/Jic.2024.9180037

Hayat, A. and Morgado-Dias, F. (2022), Deep learning-based automatic safety helmet detection system for construction safety, Applied Sciences, Vol. 12, No. 16, 8268.

10.3390/app12168268

Jocher, G., Chaurasia, A., and Qiu, J. (2023), Ultralytics YOLOv8, Retrieved from https://github.com/ultralytics/ultralytics

Kim, K., Kim, K. and Jeong, S. (2023), Application of YOLO v5 and v8 for recognition of safety risk factors at construction sites, Sustainability, Vol. 15, No. 20, 15179.

10.3390/su152015179

LOTTE E&C (2023), Lotte E&C applies industry-first AI-based technology to detect voids behind earth-retaining structures, https://www.lottecon.co.kr/Medias/notice_view?noticetypecd=003&maxdisp=10&pageno=1&searchtype=&keyword=&noticeSeq=1194 (Accessed on April 03, 2026) (In Korean).

Mien, T. L., Tu, N. D. and Van Lam, N. (2025), Deploying YOLOv8 for Real-Time Road Crack Detection on Smart Road Length Measurement Devices, Journal of Future Artificial Intelligence and Technologies, Vol. 2, No. 1, pp. 135~144.

10.62411/faith.3048-3719-102

Ministry of Land, Infrastructure and Transport (2021), Maintenance Manual for Ground Anchor (In Korean).

Park, L. G., Kim, T. O., Seo, Y. H., An, M. H., Yu, S. H. and Hong, Y. S. (2019), Current status and collapse cases of temporary earth retaining structures, Geotechnical Engineering, Vol. 35, No. 4, pp. 50~67 (In Korean).

Park, M., Bak, J. and Park, S. (2023), Small and overlapping worker detection at construction sites, Automation in Construction, 151, 104856.

10.1016/j.autcon.2023.104856

Seo, S., Chen, D., Kim, K., Kang, K., Koo, D., Chae, M., and Park, H. K. (2022), Temporary traffic control device detection for road construction projects using deep learning application, In Construction Research Congress 2022, pp. 392~401.

10.1061/9780784483961.042

Thakur, D., Pal, P., Jadhav, A., Kable, N. and Deshpande, S. (2023), Yolov8-based helmet and vest detection system for safety assessment, In 2023 International Conference on Network, Multimedia and Information Technology (NMITCON), pp. 1~10.

10.1109/NMITCON58196.2023.10275958

Varghese, R. and Sambath, M. (2024), Yolov8: A novel object detection algorithm with enhanced performance and robustness, In 2024 International conference on advances in data engineering and intelligent computing systems (ADICS) pp. 1~6.

10.1109/ADICS58448.2024.10533619

Zhang, S. and Zhang, L. (2022), Construction site safety monitoring and excavator activity analysis system, Construction Robotics, Vol. 6, No. 2, pp.151~161.

10.1007/s41693-022-00077-0

Zhang, Y., Guan, D., Zhang, S., Su, J., Han, Y. and Liu, J. (2024), GSO-YOLO: Global stability optimization YOLO for construction site detection, arXiv preprint arXiv:2407.00906.

Zhexebay, D., Skabylov, A., Turlykozhayeva, D., Bolysbay, A., Turmaganbet, U., Namazbayev, T. and Tao, M. (2025), Real-time small object detection with YOLOv8n/8s and YOLOv11n/11s models in complex natural landscapes, Physical Sciences and Technology, Vol. 12, No. 1-2, pp. 142~153.

10.26577/phst202512114

Information

Publisher :Korean Geo-Environmental Society
Publisher(Ko) :한국지반환경공학회
Journal Title :Journal of the Korean Geo-Environmental Society
Journal Title(Ko) :한국지반환경공학회 논문집
Volume : 27
No :5
Pages :37-43
Received Date : 2026-04-08
Revised Date : 2026-04-13
Accepted Date : 2026-04-21
DOI :https://doi.org/10.14481/jkges.2026.27.5.37

Journal of the Korean Geo-Environmental Society ISSN:1598-0820(Print) 2714-1233(Online) 한국지반환경공학회 논문집

All Issue