한국건설관리학회논문집 (Korean Journal of Construction Engineering and Management)

  • 제25권3호
  • /
  • Pages.58-67
  • /
  • 2024
  • /
  • 2005-6095(pISSN)
  • /
  • 2465-9703(eISSN)
한국건설관리학회 (Korea Institute of Construction Engineering and Management)
권호 표지
DOI QR코드

DOI QR Code

표준시방서 기반의 의미론적 분석을 반영한 건설 현장 사진 자동 분류 모델 개발

Development of an Automatic Classification Model for Construction Site Photos with Semantic Analysis based on Korean Construction Specification

  • 투고 : 2023.11.22
  • 심사 : 2024.02.28
  • 발행 : 2024.05.31
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

4차 산업 시대에서의 데이터는 산업의 생산성을 높이는 데 매우 중요한 역할을 하고 있다. 활용 가능한 데이터가 부족한 건설산업의 디지털화 수준을 높이기 위해서 본 연구에서는 건설 현장 사진을 공종별로 분류하는 모델을 연구하였다. 이미지만을 가지고 분류하는 기존의 이미지 분류 모델과 달리, 본 연구는 표준시방서에서 객체와 공종 간의 중요도를 추출하여 이를 분류 과정에 반영하는 방식으로 공종에 대한 의미론적인 분석을 포함한 분류 모델을 제안하였다. 객체와 공종 간의 중요도는 사진 내에서 탐지한 객체와 표준시방서의 정보를 연결하여 추출한 후 모델에 반영하였고, 이러한 방식으로 개발된 모델을 분류 프로그램에 적용하여 실제 실무에서의 유용성을 확인해 보았다. 제안한 모델은 결과에 해석가능성과 신뢰도를 높여주는 것뿐만 아니라 현장 기사들이 사진을 분류하는데 용이성을 주게 되며, 이러한 연구의 결과는 건설산업의 디지털화에 기여할 수 있을 것이다.

In the era of the fourth industrial revolution, data plays a vital role in enhancing the productivity of industries. To advance digitalization in the construction industry, which suffers from a lack of available data, this study proposes a model that classifies construction site photos by work types. Unlike traditional image classification models that solely rely on visual data, the model in this study includes semantic analysis of construction work types. This is achieved by extracting the significance of relationships between objects and work types from the standard construction specification. These relationships are then used to enhance the classification process by correlating them with objects detected in photos. This model improves the interpretability and reliability of classification results, offering convenience to field operators in photo categorization tasks. Additionally, the model's practical utility has been validated through integration into a classification program. As a result, this study is expected to contribute to the digitalization of the construction industry.

키워드

과제정보

이 논문은 정부(과학기술정보통신부)의 재원으로 한국연구재단의 지원을 받아 수행된 연구임(No. 2021R1F1A1063286).

참고문헌

  1. Akanbi, T., and Zhang, J. (2021). "Design information extraction from construction specifications to support cost estimation." Automation in Construction, 131, 103835.
  2. Castelvecchi, D. (2016). "Can we open the black box of AI?" Nature News, 538(7623), pp. 20-23. https://doi.org/10.1038/538020a
  3. Cheng, J.C., and Wang, M. (2018). "Automated detection of sewer pipe defects in closed-circuit television images using deep learning techniques." Automation in Construction, 95, pp. 155-171. https://doi.org/10.1016/j.autcon.2018.08.006
  4. Gil, D.Y. (2019). "Automated method to classify construction site photos based on deep learning." Master Dissertation, Yonsei University.
  5. Gu, H.M., Seo, J.H., and Choo, S.Y. (2019). "A Development of Facade Dataset Construction Technology Using Deep Learning-based Automatic Image Labeling." Journal of the architectural institute of Korea planning & design, 35(12), pp. 43-53.
  6. Hwang, J., Kim, J., Chi, S., and Seo, J. (2019). "Automated Training Database Development through Image Web Crawling for Construction Site Monitoring." Journal of Civil and Environmental Engineering Research, KSCE, 39(6), pp. 887-892.
  7. Kim, H., Ahn, J., Lee, T.Y., and Choi, B. (2023a). "The Object Detector for Aerial Image Using High Resolution Feature Extractor and Attention Module." The Journal of Korean Institute of Communications and Information Sciences, 48(1), pp. 1-11. https://doi.org/10.7840/kics.2023.48.1.1
  8. Kim, K., Kim, K., and Jeong, S. (2023b). "Application of YOLO v5 and v8 for Recognition of Safety Risk Factors at Construction Sites." Sustainability, 15(20), 15179.
  9. Kolar, Z., Chen, H., and Luo, X. (2018). "Transfer learning and deep convolutional neural networks for safety guardrail detection in 2D images." Automation in Construction, 89, pp. 58-70. https://doi.org/10.1016/j.autcon.2018.01.003
  10. Luo, X., Li, H., Cao, D., Dai, F., Seo, J., and Lee, S. (2018). "Recognizing diverse construction activities in site images via relevance networks of construction-related objects detected by convolutional neural networks." J. Comput. Civ. Eng, 32(3), 04018012.
  11. Mesfin, W.M., Cho, S., Lee, J., Kim, H.K., and Kim, T. (2021). "Deep-learning-based segmentation of fresh or young concrete sections from images of construction sites." Materials, 14(21), 6311.
  12. MOLIT. (2018). Smart Construction Technology Roadmap, Ministry of Land, Infrastructure (MOLIT).
  13. Moon, S., Lee, G., and Chi, S. (2022). "Automated system for construction specification review using natural language processing." Advanced Engineering Informatics, 51, 101495.
  14. Nath, N.D., Behzadan, A.H., and Paal, S.G. (2020). "Deep learning for site safety: Real-time detection of personal protective equipment." Automation in Construction, 112, 103085.
  15. Park, D.S., and Kim, H.J. (2018). "A Proposal of Join Vector for Semantic Factor Reflection in TF-IDF Based Keyword Extraction." The Journal of Korean Institute of Information Technology, 16(2), pp. 1-16.
  16. 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, 2016, pp. 779-788.
  17. RICON FOCUS. (2020). "Innovation plan for the construction industry according to the acceleration of the 'digital economy'." Korea Research Institute For Construction Policy (RICON).
  18. Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., and Chen, L.C. (2018). "Mobilenetv2: Inverted residuals and linear bottlenecks." In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 4510-4520.
  19. Seong, Y., and Yoo, W. (2022). "Korean construction industry productivity analysis." Construction & Economy Research Institute of Korea.
  20. Shorten, C., and Khoshgoftaar, T.M. (2019). "A survey on image data augmentation for deep learning." Journal of big data, 6(1), pp. 1-48. https://doi.org/10.1186/s40537-019-0197-0
  21. Siddula, M., Dai, F., Ye, Y., and Fan, J. (2016). "Unsupervised feature learning for objects of interest detection in cluttered construction roof site images." Procedia Engineering, 145, pp. 428-435. https://doi.org/10.1016/j.proeng.2016.04.010
  22. Sokolova, M., and Lapalme, G. (2009). "A systematic analysis of performance measures for classification tasks." Information processing & management, 45(4), pp. 427-437. https://doi.org/10.1016/j.ipm.2009.03.002
  23. Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., and Salakhutdinov, R. (2014). "Dropout: a simple way to prevent neural networks from overfitting." The Journal of Machine Learning Research, 15(1), pp. 1929-1958.
  24. Voulodimos, A., Doulamis, N., Doulamis, A., and Protopapadakis, E. (2018). "Deep learning for computer vision: A brief review." Computational intelligence and neuroscience, 2018.
  25. Wang, Z., Kim, J.S., Ham, N.H., and Kim, J.J. (2022). "Proposal and Verification of the Faster R-CNN Regarding the Worker and Machine Interference Scope Detection Model to Prevent On-site Safety Accidents." Journal of the Architectural Institute of Korea, 38(4), pp. 217-228.
  26. Wang, Z., Wu, Y., Yang, L., Thirunavukarasu, A., Evison, C., and Zhao, Y. (2021). "Fast personal protective equipment detection for real construction sites using deep learning approaches." Sensors, 21(10), 3478.
  27. Xu, N., Zhou, X., Guo, C., Xiao, B., Wei, F., and Hu, Y. (2022). "Text Mining Applications in the Construction Industry: Current Status, Research Gaps, and Prospects." Sustainability, 14(24), 16846.
  28. Zhao, Z., Fan, X., Xu, G., Zhang, L., Qi, Y., and Zhang, K. (2017). "Aggregating deep convolutional feature maps for insulator detection in infrared images." IEEE Access, 5, pp. 21831-21839. https://doi.org/10.1109/ACCESS.2017.2757030