딥러닝기반 초해상화 영상 복원을 이용한 콘크리트 균열 탐지의 성능 평가
Performance evaluation of concrete crack detection using deep learning-based super-resolution image reconstruction
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
콘크리트 터널과 같은 지하 구조물의 안전한 유지관리는 균열과 같은 열화 현상을 조기에 탐지하는 데 필수적이다. 노후구조물이 증가함에 따라, 효율적인 안전 관리를 위해 다양한 컴퓨터비전 기술이 구조물 유지 관리에 활용되고 있다. 그러나 현장에서 수집된 영상은 여러 이유로 저해상도 및 저품질로 수집될 수 있으며, 이는 특히 딥러닝을 활용한 균열 탐지에서 정확도 저하로 이어질 수 있다. 따라서, 본 연구에서는 초해상화 알고리즘을 적용하여 저해상도 영상에서도 높은 탐지 성능을 유지할 수 있는 방법을 제안하였다. 다양한 비율의 저해상도 영상을 대상으로 초해상화 알고리즘을 적용하여균열을 효과적으로 탐지하는 방안을 제시하고, 이를 통해 여러 초해상화 알고리즘의 효용성을 검토하였다. 실험을 통해제안된 방법이 다양한 입력 데이터 조건에서도 안정적이고 향상된 균열 탐지 성능을 나타냄을 확인하였다. 본 연구는 터널 유지 관리 측면에서 초해상화 알고리즘의 활용 가능성을 보여주며, 향후 실제 터널 유지관리에서 정확한 관리에 기여할 것으로 기대된다
Abstract
The maintenance of underground structures, such as concrete tunnels, is essential forthe early detection of deterioration phenomena, such as cracks. As the number of agingstructures increases, various computer-vision technologies are being applied to efficientlymaintain these aging structures. However, images collected in the field can oftenbe low-resolution and low-quality due to various reasons, which can lead to reducedaccuracy, especially in crack detection using deep learning. Therefore, this study proposesa method to maintain high detection performance even in low-resolution imagesby applying super-resolution algorithms. By applying super-resolution algorithms toimages with varying proportions of low-quality images, we present an effective methodfor crack detection and evaluate the utility of several super-resolution algorithms. Theexperimental results confirmed that the proposed method exhibits stable and improvedcrack detection performance across various input data conditions. This study demonstratesthe potential of super-resolution algorithms in tunnel maintenance and is expectedto contribute to accurate maintenance in future real-world tunnel management.
참고문헌 (27)
- Cha, Y. J., "Deep learning-based crack damage detection using convolutional neural networks", Computer-Aided Civil and Infrastructure Engineering, 32(5), pp.361~378 (2017)
- Croitoru, F. A., "Diffusion models in vision : a survey", IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(9), pp.10850~10869 (2023)
- Ding, K., "Image quality assessment : unifying structure and texture similarity", IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(5), pp.2567~2581 (2022)
- Dong, C., "Image super-resolution using deep convolutional networks", IEEE Transactions on Pattern Analysis and Machine Intelligence, 38(2), pp.295~307 (2016)
- Gilman, A., "Interpolation models for image super-resolution", Proceedings of the 4th IEEE International Symposium on Electronic Design, Test and Applications, pp.55~60 (2008)
- Kim, J., "Lightweight pixel-wise segmentation for efficient concrete crack detection using hierarchical convolutional neural network", Smart Materials and Structures, 30(4), pp.045023~ (2021)
- Kulkarni, S., "Proceedings of the Computer Vision - ECCV 2022 Workshops", pp.179~195 (2022)
- Ledig, C., "Photo-realistic single image super-resolution using a generative adversarial network", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.4681~4690 (2017)
- Liang, J., "Swinir : Image restoration using swin transformer", Proceedings of the IEEE, 2021, 1833~1844
- Liu, Y., "DeepCrack : a deep hierarchical feature learning architecture for crack segmentation", Neurocomputing, 338, pp.139~153 (2019)
- Long, J., "Fully convolutional networks for semantic segmentation", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.3431~3440 (2015)
- Lu, Z., "Transformer for single image super-resolution", Proceedings of the IEEE, 2022, 457~466
- Maguire, M., "SDNET2018: a concrete crack image dataset for machine learning applications", Utah State University, ~ (2018)
- MOLIT, "Detailed guidelines for the safety control and maintenance of establishments", Ministry of Land, Infrastructure and Transport, Korea Authority of Land & Infrastructure Safety, ~ (2023)
- Ozgenel, F., "Concrete crack images for classification", Mendeley data-v1, ~ (2018)
- Pan, Y., "A spatial-channel hierarchical deep learning network for pixellevel automated crack detection", Automation in Construction, 119, pp.103357~ (2020)
- Ren, Y., "Image-based concrete crack detection in tunnels using deep fully convolutional networks", Construction and Building Materials, 234, pp.117367~ (2020)
- Romera, E., "ERFNet : efficient residual factorized ConvNet for real-time semantic segmentation", IEEE Transactions on Intelligent Transportation Systems, 19(1), pp.263~272 (2018)
- Saharia, C., "Image super-resolution via iterative refinement", IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(4), pp.4713~4726 (2023)
- Sheikh, H. R., "Image information and visual quality", IEEE Transactions on Image Processing, 15(2), pp.430~444 (2006)
- Wang, Y., "LEDNet: a lightweight encoder-decoder network for real-time semantic segmentation", Proceedings of the 2019 IEEE International Conference on Image Processing (ICIP), pp.1860~1864 (2019)
- Wang, Z., "Image quality assessment : from error visibility to structural similarity", IEEE Transactions on Image Processing, 13(4), pp.600~612 (2004)
- Wang, Z., "Deep learning for image super-resolution : a survey", IEEE Transactions on Pattern Analysis and Machine Intelligence, 43(10), pp.3365~3387 (2021)
- Wu, T., "CGNet : A light-weight context guided network for semantic segmentation", IEEE Transactions on Image Processing, 30, pp.1169~1179 (2021)
- Yamaguchi, T., "Fast crack detection method for large-size concrete surface images using percolation-based image processing", Machine Vision and Applications, 21(5), pp.797~809 (2010)
- Yu, Z., "Cracklab : a high-precision and efficient concrete crack segmentation and quantification network", Developments in the Built Environment, 12, pp.100088~ (2022)
- Zhang, R., "The unreasonable effectiveness of deep features as a perceptual metric", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp.586~595 (2018)