한국전산구조공학회논문집 (Journal of the Computational Structural Engineering Institute of Korea)

  • 제36권5호
  • /
  • Pages.323-330
  • /
  • 2023
  • /
  • 1229-3059(pISSN)
  • /
  • 2287-2302(eISSN)
한국전산구조공학회 (Computational Structural Engineering Institute of Korea)
권호 표지
DOI QR코드

DOI QR Code

U-net 딥러닝 기법을 활용한 PVA 섬유 보강 시멘트 복합체의 섬유 분리

Phase Segmentation of PVA Fiber-Reinforced Cementitious Composites Using U-net Deep Learning Approach

  • 서지우 (연세대학교 건설환경공학과) ;
  • 한동석 (연세대학교 건설환경공학과)
  • Jeewoo Suh (Department of Civil and Environmental Engineering, Yonsei University) ;
  • Tong-Seok Han (Department of Civil and Environmental Engineering, Yonsei University)
  • 투고 : 2023.06.30
  • 심사 : 2023.08.21
  • 발행 : 2023.10.31
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

PVA 섬유 보강 시멘트 복합체는 매우 복잡한 미세구조를 가지고 있으며, 재료의 거동을 정확히 평가하기 위해서는 미세구조 특성을 반영하여 실제 실험과 시너지효과를 내며 효율적인 재료 설계를 가능하게 하는 해석 모델의 개발이 중요하다. PVA 섬유 보강 시멘트 복합체의 역학적 성능은 PVA 섬유의 방향성에 큰 영향을 받는다. 그러나 마이크로-CT 이미지로부터 얻은 PVA 섬유의 회색조 값을 인접한 상과 구분하기 어려워, 섬유 분리 과정에 많은 시간이 소요된다. 본 연구에서는 섬유의 3차원 분포를 얻기 위하여 0.65㎛3의 복셀 크기를 가지는 마이크로-CT 이미지 촬영을 수행하였다. 학습에 사용될 학습 데이터를 생성하기 위해 히스토그램, 형상, 그리고 구배 기반 상 분리 방법을 적용하였다. 본 연구에서 제안된 U-net 모델을 활용하여 PVA 섬유 보강 시멘트 복합체의 마이크로- CT 이미지로부터 섬유를 분리하는 학습을 수행하였다. 훈련의 정확도를 높이기 위해 데이터 증강을 적용하였으며, 총 1024개의 이미지를 훈련 데이터로 사용하였다. 모델의 성능은 정확도, 정밀도, 재현율, F1 스코어를 평가하였으며, 학습된 모델의 섬유 분리 성능이 매우 높고 효율적이며, 다른 시편에도 적용될 수 있음을 확인하였다.

The development of an analysis model that reflects the microstructure characteristics of polyvinyl alcohol (PVA) fiber-reinforced cementitious composites, which have a highly complex microstructure, enables synergy between efficient material design and real experiments. PVA fiber orientations are an important factor that influences the mechanical behavior of PVA fiber-reinforced cementitious composites. Owing to the difficulty in distinguishing the gray level value obtained from micro-CT images of PVA fibers from adjacent phases, fiber segmentation is time-consuming work. In this study, a micro-CT test with a voxel size of 0.65 ㎛3 was performed to investigate the three-dimensional distribution of fibers. To segment the fibers and generate training data, histogram, morphology, and gradient-based phase-segmentation methods were used. A U-net model was proposed to segment fibers from micro-CT images of PVA fiber-reinforced cementitious composites. Data augmentation was applied to increase the accuracy of the training, using a total of 1024 images as training data. The performance of the model was evaluated using accuracy, precision, recall, and F1 score. The trained model achieved a high fiber segmentation performance and efficiency, and the approach can be applied to other specimens as well.

키워드

과제정보

본 연구는 2022년도 해양수산부 재원으로 해양수산과학진흥원의 지원을 받아 수행된 연구임(G22202204022201).

참고문헌

  1. Abrishambaf, A., Pimentel, M., Nunes, S. (2017) Influence of Fibre Orientation on the Tensile Behaviour of Ultra-High Performance Fibre Reinforced Cementitious Composites, Cem. & Concr. Res., 97, pp.28~40. https://doi.org/10.1016/j.cemconres.2017.03.007
  2. Bangaru, S.S., Wang, C., Zhou, X., Hassan, M. (2022) Scanning Electron Microscopy (SEM) Image Segmentation for Microstructure Analysis of Concrete using U-net Convolutional Neural Network, Autom. Constr., 144, p.104602.
  3. Chen, Z., Ting, D., Newbury, R., Chen, C. (2021) Semantic Segmentation for Partially Occluded Apple Trees based on Deep Learning, Comput. & Electron. Agric., 181, p.105952.
  4. Chicco, D., Jurman, G. (2020) The Advantages of the Matthews Correlation Coefficient (MCC) over F1 Score and Accuracy in Binary Classification Evaluation. BMC Genom., 21, pp. 1~13. https://doi.org/10.1186/s12864-019-6419-1
  5. Chung, S.Y., Kim, J.S., Stephan, D., Han, T.S. (2019) Overview of the use of Micro-Computed Tomography(Micro-CT) to Investigate the Relation between the Material Characteristics and Properties of Cement-based Materials, Constr. & Build. Mater., 229, p.116843.
  6. Gonzalez, R. (2010) Digital Image Processing using MATLAB (R), India: McGraw Hill Education India.
  7. Goutte, C., Gaussier, E. (2005) A Probabilistic Interpretation of Precision, Recall and F-Score, with Implication for Evaluation, In: Advances in Information Retrieval: 27th European Conference on IR Research, ECIR 2005, pp.345~359.
  8. Hao, Y., Cheng, L., Hao, H., Shahin, M.A. (2018) Enhancing Fiber/Matrix Bonding in Polypropylene Fiber Reinforced Cementitious Composites by Microbially Induced Calcite Precipitation Pre-treatment, Cem. & Concr. Compos., 88, pp. 1~7. https://doi.org/10.1016/j.cemconcomp.2018.01.001
  9. Hong, L., Zhang, P., Liu, D., Gao, P., Zhan, B., Yu, Q., Sun, L. (2021) Effective Segmentation of Short Fibers in Glass Fiber Reinforced Concrete's X-ray Images using Deep Learning Technology, Mater.& Des., 210, p.110024.
  10. Kang, S.T., Kim, J.K. (2011) The Relation between Fiber Orientation and Tensile behavior in an Ultra High Performance Fiber Reinforced Cementitious Composites (UHPFRCC), Cem. & Concr. Res., 41(10), pp.1001~1014. https://doi.org/10.1016/j.cemconres.2011.05.009
  11. Kim, J.H., Robertson, R.E. (1998) Effects of Polyvinyl Alcohol on Aggregate-Paste Bond Strength and the Interfacial Transition Zone, Adv. Cem. Based Mater., 8(2), pp.66~76. https://doi.org/10.1016/S1065-7355(98)00009-1
  12. Kim, J.S., Lim, J.H., Stephan, D., Park, K., Han, T.S. (2022a) Mechanical behavior Comparison of Single and Multiple Phase Models for Cement Paste using Micro-CT Images and Nanoindentation, Constr. & Build. Mater., 342, p.127938.
  13. Kim, J.S., Suh, J., Pae, J., Moon, J., Han, T.S. (2022b) Gradient-based Phase Segmentation Method for Characterization of Hydrating Cement Paste Microstructures Obtained from X-ray Micro-CT, J. Build. Eng., 46, p.103721.
  14. Liu, J., Li, C., Liu, J., Cui, G., Yang, Z. (2013) Study on 3D Spatial Distribution of Steel Fibers in Fiber Reinforced Cementitious Composites through Micro-CT Technique, Constr. & Build. Mater., 48, pp.656~661. https://doi.org/10.1016/j.conbuildmat.2013.07.052
  15. Lorenzoni, R., Curosu, I., Paciornik, S., Mechtcherine, V., Oppermann, M., Silva, F. (2020) Semantic Segmentation of the Micro-Structure of Strain-Hardening Cement-based Composites (SHCC) by Applying Deep Learning on Micro-Computed Tomography Scans, Cem. & Concr. Compos., 108, p.103551.
  16. Lu, M., Xiao, H., Liu, M., Feng, J. (2023) Carbon Fiber Surface Nano-Modification and Enhanced Mechanical Properties of Fiber Reinforced Cementitious Composites, Constr. & Build. Mater., 370, p.130701.
  17. Meng, D., Huang, T., Zhang, Y.X., Lee, C.K. (2017) Mechanical behaviour of a Polyvinyl Alcohol Fibre Reinforced Engineered Cementitious Composite (PVA-ECC) using Local Ingredients, Constr. & Build. Mater., 141, pp.259~270. https://doi.org/10.1016/j.conbuildmat.2017.02.158
  18. Miletic, M., Kumar, L.M., Arns, J.Y., Agarwal, A., Foster, S.J., Arns, C., Peric, D. (2020) Gradient-based Fibre Detection Method on 3D Micro-CT Tomographic Image for Defining Fibre Orientation Bias in Ultra-High-Performance Concrete, Cem. & Concr. Res., 129, p.105962.
  19. Ronneberger, O., Fischer, P., Brox, T. (2015) U-net: Convolutional Networks for Biomedical Image Segmentation, In Medical Image Computing and Computer-Assisted Intervention-MICCAI 2015: 18th International Conference, pp.234~241.
  20. Sadrmomtazi, A., Tahmouresi, B., Saradar, A. (2018) Effects of Silica Fume on Mechanical Strength and Microstructure of Basalt Fiber Reinforced Cementitious Composites (BFRCC), Constr. & Build. Mater., 162, pp.321~333. https://doi.org/10.1016/j.conbuildmat.2017.11.159
  21. Singh, N.B., Rai, S. (2001) Effect of Polyvinyl Alcohol on the Hydration of Cement with Rice Husk Ash, Cem. & Concr. Res., 31(2), pp.239~243. https://doi.org/10.1016/S0008-8846(00)00475-0
  22. Skarzynski, L., Suchorzewski, J. (2018) Mechanical and Fracture Properties of Concrete Reinforced with Recycled and Industrial Steel Fibers using Digital Image Correlation Technique and X-ray Micro Computed Tomography, Constr. & Build. Mater., 183, pp.283~299. https://doi.org/10.1016/j.conbuildmat.2018.06.182
  23. Tao, J., Gong, H., Wang, F., Luo, X., Qiu, X., Liu, J. (2022) Deep Learning based Automated Segmentation of Air-Void System in Hardened Concrete Surface using Three Dimensional Reconstructed Images, Constr. & Build. Mater., 324, p.126717.
  24. Thong, C.C., Teo, D.C.L., Ng, C.K. (2016) Application of Polyvinyl Alcohol (PVA) in Cement-based Composite Materials: A Review of Its Engineering Properties and Microstructure Behavior, Constr. & Build. Mater., 107, pp.172~180. https://doi.org/10.1016/j.conbuildmat.2015.12.188
  25. Xie, C., Cao, M., Yin, H., Guan, J., Wang, L. (2021) Effects of Freeze-Thaw Damage on Fracture Properties and Microstructure of Hybrid Fibers Reinforced Cementitious Composites Containing Calcium Carbonate Whisker, Constr. & Build. Mater., 300, p.123872.
  26. Xu, K., Jin, Q., Li, J., Ushizima, D.M., Li, V.C., Kurtis, K.E., Monteiro, P.J. (2023) In-Situ Microtomography Image Segmentation for Characterizing Strain-Hardening Cementitious Composites under Tension using Machine Learning, Cem. & Concr. Res., 169, p.107164.
  27. Yang, Z.J., Qsymah, A., Peng, Y.Z., Margetts, L., Sharma, R. (2020) 4D Characterisation of Damage and Fracture Mechanisms of Ultra High Performance Fibre Reinforced Concrete by In-situ Micro X-Ray Computed Tomography Tests, Cem. & Concr. Compos., 106, p.103473.
  28. Ye, Z.B., Huang, R.Y., Li, Y.C., Lv, L., Zhao, K., Zhang, Y.L., Ma, J., Lin, J.J. (2018) Steel Fiber-Reinforced Concrete under Impact Loading Dynamic Constitutive Equation, Constr. & Build. Mater., 190, pp.1049~1055. https://doi.org/10.1016/j.conbuildmat.2018.09.118