Abstract
This paper aims to establish an automatic and accurate pore identification method for pervious concrete. The residual module and mixed loss functions were introduced to the original UNet network to obtain the improved UNet. CT scanning was conducted on the six groups of pervious concrete samples with different aggregate sizes to obtain the initial dataset. The initial dataset was marked and enhanced, and then the pore recognition model was trained. The influence of image brightness and contrast on pore identification was analyzed. The fusion algorithm was used to improve the robustness of the model. The results show that during model training, R-UNet began to converge 20 epochs earlier than UNet and the loss value was smaller. Moreover, the maximum increase of mIoU and mDice was 10.3% and 11.7% respectively, and the maximum decrease of mHD was 14.1%. The fusion algorithm could improve the segmentation accuracy of pores in brightness anomaly images. Compared with threshold segmentation method, the method proposed in this paper could improve the accuracy of pore edge segmentation and the “fine pores” identification, and reduced the pore identification defects. The value of mHD was decreased by 48.7%–72.4%, and the efficiency of pore identification was greatly improved.
Similar content being viewed by others
References
Chen Y, Wang K, Wang X (2013) Strength, fracture and fatigue of pervious Concrete. Construction & Building Materials 42(5):97–104, DOI: https://doi.org/10.1016/j.conbuildmat.2013.01.006
Chindaprasirt P, Hatanaka S, Chareerat T (2006) Cement paste characteristics and porous concrete properties. Construction & Building Materials 22(5):894–901, DOI: https://doi.org/10.1016/j.conbuildmat.2006.12.007
Cubuk ED, Zoph B, Mane D (2019) AutoAugment: Learning augmentation strategies from data. International Conference on Computer Vision and Pattern Detection (CVPR), 113–123, DOI: https://doi.org/10.1109/CVPR.2019.00020
Deo O, Neithalath N (2010) Compressive behavior of pervious concretes and a quantification of the influence of random pore structure features. Materials Science & Engineering A 528(1):402–412, DOI: https://doi.org/10.1016/j.msea.2010.09.024
Güneyisi E, Gesoğlu M, Kareem Q (2016) Effect of different substitution of natural aggregate by recycled aggregate on performance characteristics of pervious concrete. Material & Structures 49(1–2):521–536, DOI: https://doi.org/10.1617/s11527-014-0517-y
He KM, Zhang XY, Ren SQ, Sun J (2016) Deep residual learning for image recognition. CoRR, 2015, abs/1512.03385, DOI: https://doi.org/10.1109/CVPR.2016.90
Hussain Z, Gimenez F, Yi D (2019) Differential data augmentation techniques for medical imaging classification tasks. AMIA Annual Symposium Proceedings, 979–984
Kayhanian M, Anderson D, Harvey JT (2012) Permeability measurement and scan imaging to assess clogging of pervious concrete pavements in parking lots. Journal of Environmental Management 95(1):114–123, DOI: https://doi.org/10.1016/j.jenvman.2011.09.021
Li DH (2019) Research on pavement distresses location and measurement under the heavy traffic using deep learning & UAV. MSc Thesis, Chang’an University, Xian, China (in Chinese)
Liu Z, Cao Y, Wang Y, Wang W (2019) Computer vision-based concrete crack detection using U-net fully convolutional networks. Automation in Construction 104:129–139, DOI: https://doi.org/10.1016/j.autcon.2019.04.005
Liu GX, Yang WH (2001) Image fusion scheme of pixel-level and multi-operator for infra-red and visible light images. Journal of Infrared and Millimeter Waves (3):207–210, DOI: https://doi.org/10.3321/j.issn:1001-9014.2001.03.011
Milletari F, Navab N, Ahmadi S-A (2016) V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation. CoRR, abs/1606.04797, DOI: https://doi.org/10.1109/3DV.2016.79
Oinam Y, Ju S, Gwon S (2022) Characteristics of GGBFS-based pervious concrete considering rheological properties of the binder. International Journal of Concrete Structures and Materials 16:62, DOI: https://doi.org/10.1186/s40069-022-00551-7
Park S, Jua S, Kim HK (2022) Effect of the rheological properties of fresh binder on the compressive strength of pervious concrete. Journal of Materials Research and Technology 17:636–648, DOI: https://doi.org/10.1016/j.jmrt.2022.01.045
Park SB, Seo DS, Lee J (2005) Studies on the sound absorption characteristics of porous concrete based on the content of recycled aggregate and target void ratio. Cement & Concrete Research 35(9):1846–1854, DOI: https://doi.org/10.1016/j.cemconres.2004.12.009
Radlinska A, Welker A, Greising K (2012) Long-term field performance of pervious concrete pavement. Advances in Civil Engineering 65–73, DOI: https://doi.org/10.1155/2012/380795
Sheng W (2019) Research and application of fast image analysis pore structure of concrete based on deep learning. East China Jiaotong University, Jiangxi, China (in Chinese)
Sumanasooriya MS, Neithalath N (2011) Pore structure features of pervious concretes proportioned for desired porosities and their performance prediction. Cement and Concrete Composites 33(8): 778–787, DOI: https://doi.org/10.1016/j.cemconcomp.2011.06.002
Tang HY, Yan SM (2020) Research on the influence of different aggregate particle size on the performance of permeable concrete. Highway Engineering 45(4):193–196, DOI: https://doi.org/10.19782/j.cnki.1674-0610.2020.04.034
Xing YY, Yang F, Tang YJ (2020) Ultrasound fetal head edge detection using fusion UNet++. Journal of Image and Graphics 25(2):0366–0377
Yang Q, Shi W, Chen J, Lin W (2020) Deep convolution neural network-based transfer learning method for civil infrastructure crack detection. Automation in Construction 116:103199, DOI: https://doi.org/10.1016/j.autcon.2020.103199
Yu F, Sun DQ, Wang J, Hu MJ (2019a) Influence of aggregate size on compressive strength of pervious concrete. Construction & Building Materials 209(2):463–475, DOI: https://doi.org/10.1016/j.conbuildmat.2019.03.140
Yu X, Wu X, Luo C (2019b) Deep learning in remote sensing scene classification: A data augmentation enhanced convolutional neural network framework. GIS Science and Remote Sensing 54(5):741–758, DOI: https://doi.org/10.1080/15481603.2017.1323377
Zhang J, Ma GD, Ming RP (2017) Numerical study on seepage flow in pervious concrete based on 3d CT imaging. Construction & Building Materials 161:468–478, DOI: https://doi.org/10.1016/j.conbuildmat.2017.11.149
Zhang H, Zhang R, Sun D (2022) Analyzing the pore structure of pervious concrete based on the deep learning framework of mask R-CNN. Construction & Building Materials 318(2), DOI: https://doi.org/10.1016//j.conbuildmat.2021.125987
Zhong R, Wille K (2016) Linking pore system characteristics to the compressive behavior of pervious concrete. Cement & Concrete Composites 70(2):130–138, DOI: https://doi.org/10.1016/j.cemconcomp.2016.03.016
Zhou ZS, Chen XM, Zhang HY, Wan HL, Zhao JY, Zhang T, Wang XY (2022) Application of improved UNet network in the recognition and segmentation of hemorrhage regions in brain CT images. Journal of Sichuan University. Medical Science Edition 53(1)
Zhou S, Sheng W, Wang Z (2019) Quick image analysis of concrete pore structure based on deep learning. Construction and Building Materials 208(3):144–157, DOI: https://doi.org/10.1016/j.conbuildmat.2019.03.006
Acknowledgments
This work was supported by National Natural Science Foundation of China (Grant No.61871258), Research Fund for Excellent Dissertation of China Three Gorges University (Grant No. 2021BSPY005), Natural Science Research Project of Yichang (A22-3-003), and Qinghai Province Science and Technology Plan Project (No:2021-ZJ-733).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Yu, F., Li, K., Zhang, H. et al. Pore Structure Identification Method for Pervious Concrete Based on Improved UNet and Fusion Algorithm. KSCE J Civ Eng 27, 4834–4848 (2023). https://doi.org/10.1007/s12205-023-2316-x
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12205-023-2316-x