Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
139 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
46 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Real-time High-Resolution Neural Network with Semantic Guidance for Crack Segmentation (2307.00270v2)

Published 1 Jul 2023 in cs.CV

Abstract: Deep learning plays an important role in crack segmentation, but most work utilize off-the-shelf or improved models that have not been specifically developed for this task. High-resolution convolution neural networks that are sensitive to objects' location and detail help improve the performance of crack segmentation, yet conflict with real-time detection. This paper describes HrSegNet, a high-resolution network with semantic guidance specifically designed for crack segmentation, which guarantees real-time inference speed while preserving crack details. After evaluation on the composite dataset CrackSeg9k and the scenario-specific datasets Asphalt3k and Concrete3k, HrSegNet obtains state-of-the-art segmentation performance and efficiencies that far exceed those of the compared models. This approach demonstrates that there is a trade-off between high-resolution modeling and real-time detection, which fosters the use of edge devices to analyze cracks in real-world applications.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (37)
  1. Crack Segmentation on UAS-based Imagery using Transfer Learning, in: 2019 International Conference on Image and Vision Computing New Zealand (IVCNZ), pp. 1–6. doi:10.1109/IVCNZ48456.2019.8960998. iSSN: 2151-2205.
  2. Coco-stuff: Thing and stuff classes in context, in: Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 1209–1218.
  3. Automatic crack segmentation using deep high-resolution representation learning. Applied Optics 60, 6080–6090. URL: https://opg.optica.org/ao/abstract.cfm?uri=ao-60-21-6080, doi:10.1364/AO.423406. publisher: Optica Publishing Group.
  4. Rethinking Atrous Convolution for Semantic Image Segmentation. URL: http://arxiv.org/abs/1706.05587, doi:10.48550/arXiv.1706.05587. arXiv:1706.05587 [cs].
  5. The cityscapes dataset for semantic urban scene understanding, in: Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 3213–3223.
  6. Crack detection and quantification for concrete structures using UAV and transformer. Automation in Construction 152, 104929. URL: https://www.sciencedirect.com/science/article/pii/S0926580523001899, doi:https://doi.org/10.1016/j.autcon.2023.104929.
  7. Rethinking BiSeNet For Real-time Semantic Segmentation. URL: http://arxiv.org/abs/2104.13188, doi:10.48550/arXiv.2104.13188. arXiv:2104.13188 [cs].
  8. Pavement crack detection based on transformer network. Automation in Construction 145, 104646. URL: https://www.sciencedirect.com/science/article/pii/S0926580522005167, doi:10.1016/j.autcon.2022.104646.
  9. Searching for MobileNetV3. URL: http://arxiv.org/abs/1905.02244, doi:10.48550/arXiv.1905.02244. arXiv:1905.02244 [cs].
  10. Machine learning for crack detection: review and model performance comparison. Journal of Computing in Civil Engineering 34, 04020038. Publisher: American Society of Civil Engineers.
  11. Pixelwise asphalt concrete pavement crack detection via deep learning‐based semantic segmentation method. Structural Control and Health Monitoring 29. URL: https://onlinelibrary.wiley.com/doi/10.1002/stc.2974, doi:10.1002/stc.2974.
  12. An Efficient High-Resolution Global–Local Network to Detect Lunar Features for Space Energy Discovery. Remote Sensing 14, 1391. URL: https://www.mdpi.com/2072-4292/14/6/1391, doi:10.3390/rs14061391. number: 6 Publisher: Multidisciplinary Digital Publishing Institute.
  13. Two-step deep learning approach for pavement crack damage detection and segmentation. International Journal of Pavement Engineering , 1–14URL: https://www.tandfonline.com/doi/full/10.1080/10298436.2022.2065488, doi:10.1080/10298436.2022.2065488.
  14. Self-reconfigurable façade-cleaning robot equipped with deep-learning-based crack detection based on convolutional neural networks. Automation in Construction 108, 102959.
  15. CrackSeg9k: A Collection and Benchmark for Crack Segmentation Datasets and Frameworks. URL: http://arxiv.org/abs/2208.13054, doi:10.48550/arXiv.2208.13054. arXiv:2208.13054 [cs].
  16. Automatic Tunnel Crack Inspection Using an Efficient Mobile Imaging Module and a Lightweight CNN. IEEE Transactions on Intelligent Transportation Systems , 1–14doi:10.1109/TITS.2021.3138428.
  17. DeepCrack: A deep hierarchical feature learning architecture for crack segmentation. Neurocomputing 338, 139–153. URL: https://linkinghub.elsevier.com/retrieve/pii/S0925231219300566, doi:10.1016/j.neucom.2019.01.036.
  18. Image-Based Crack Detection Methods: A Review. Infrastructures 6, 115. Publisher: Multidisciplinary Digital Publishing Institute.
  19. Deep Dual-Resolution Networks for Real-Time and Accurate Semantic Segmentation of Traffic Scenes. IEEE Transactions on Intelligent Transportation Systems 24, 3448–3460. doi:10.1109/TITS.2022.3228042. conference Name: IEEE Transactions on Intelligent Transportation Systems.
  20. PP-LiteSeg: A Superior Real-Time Semantic Segmentation Model. URL: http://arxiv.org/abs/2204.02681, doi:10.48550/arXiv.2204.02681. arXiv:2204.02681 [cs].
  21. U-net: Convolutional networks for biomedical image segmentation, in: International Conference on Medical image computing and computer-assisted intervention, Springer. pp. 234–241.
  22. Performance evaluation of crack identification using density-based topology optimization for experimentally visualized ultrasonic wave propagation. Mechanics of Materials 172, 104406.
  23. DMA-Net: DeepLab With Multi-Scale Attention for Pavement Crack Segmentation. IEEE Transactions on Intelligent Transportation Systems , 1–12URL: https://ieeexplore.ieee.org/document/9741463/, doi:10.1109/TITS.2022.3158670.
  24. Semantic Segmentation Network for Surface Defect Detection of Automobile Wheel Hub Fusing High-Resolution Feature and Multi-Scale Feature. Applied Sciences 11, 10508. URL: https://www.mdpi.com/2076-3417/11/22/10508, doi:10.3390/app112210508. number: 22 Publisher: Multidisciplinary Digital Publishing Institute.
  25. Towards Interpretable Semantic Segmentation via Gradient-weighted Class Activation Mapping. Proceedings of the AAAI Conference on Artificial Intelligence 34, 13943–13944. URL: http://arxiv.org/abs/2002.11434, doi:10.1609/aaai.v34i10.7244. arXiv:2002.11434 [cs, eess].
  26. U-HRNet: Delving into Improving Semantic Representation of High Resolution Network for Dense Prediction. URL: http://arxiv.org/abs/2210.07140, doi:10.48550/arXiv.2210.07140. arXiv:2210.07140 [cs].
  27. Deep high-resolution representation learning for visual recognition. IEEE transactions on pattern analysis and machine intelligence 43, 3349–3364. Publisher: IEEE.
  28. A learning-based crack defect detection and 3d localization framework for automated fluorescent magnetic particle inspection. Expert Systems with Applications 214, 118966.
  29. An eddy current testing method based on magnetic induction intensity for detecting cracks in steel bridge decks. Journal of Performance of Constructed Facilities 37, 04023014.
  30. Pavement crack detection with hybrid-window attentive vision transformers. International Journal of Applied Earth Observation and Geoinformation 116, 103172. URL: https://www.sciencedirect.com/science/article/pii/S1569843222003600, doi:10.1016/j.jag.2022.103172.
  31. Pavement crack detection from CCD images with a locally enhanced transformer network. International Journal of Applied Earth Observation and Geoinformation 110, 102825. URL: https://www.sciencedirect.com/science/article/pii/S1569843222000279, doi:10.1016/j.jag.2022.102825.
  32. Pixel-level pavement crack detection using enhanced high-resolution semantic network. International Journal of Pavement Engineering , 1–15URL: https://www.tandfonline.com/doi/full/10.1080/10298436.2021.1985491, doi:10.1080/10298436.2021.1985491.
  33. RIIAnet: A Real-Time Segmentation Network Integrated with Multi-Type Features of Different Depths for Pavement Cracks. Applied Sciences 12, 7066. URL: https://www.mdpi.com/2076-3417/12/14/7066, doi:10.3390/app12147066.
  34. BiSeNet V2: Bilateral Network with Guided Aggregation for Real-time Semantic Segmentation. URL: http://arxiv.org/abs/2004.02147, doi:10.48550/arXiv.2004.02147. arXiv:2004.02147 [cs].
  35. Object-Contextual Representations for Semantic Segmentation, in: Vedaldi, A., Bischof, H., Brox, T., Frahm, J.M. (Eds.), Computer Vision – ECCV 2020, Springer International Publishing, Cham. pp. 173–190. URL: https://link.springer.com/10.1007/978-3-030-58539-6_11, doi:10.1007/978-3-030-58539-6_11. series Title: Lecture Notes in Computer Science.
  36. A Recurrent Adaptive Network: Balanced Learning for Road Crack Segmentation with High-Resolution Images. Remote Sensing 14, 3275. URL: https://www.mdpi.com/2072-4292/14/14/3275, doi:10.3390/rs14143275. number: 14 Publisher: Multidisciplinary Digital Publishing Institute.
  37. Pyramid scene parsing network, in: Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 2881–2890.
Citations (19)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now