Rock mass integrity grade recognition of TBM tunnel based on rock fragment image and SE-enhanced Inception-VGG19 network

Abstract

The rock mass integrity is closely related to the stability of the surrounding rock in the tunnel boring machine (TBM) tunnel, and has a crucial impact on the optimization of tunnelling control parameters and support de-cisions. The current research on the perception of surrounding rock information in TBM tunnel based on rock fragment images, only roughly identifies the rock mass grade or customized class, with little attention paid to more detailed rock mass integrity. This research proposes an accurate method for recognizing the rock mass integrity grade in TBM tunnel based on the rock fragment image and Squeeze-and-Excitation (SE)-enhanced Inception-Visual Geometry Group (VGG)19 network. Firstly, the data acquisition systems are developed and the relevant data are collected. Then, the relationship between the particle size distribution of rock fragments and the rock mass integrity is analyzed. Finally, a novel SE-enhanced Inception-VGG19 (SI-VGG19) network is de-signed and a model for recognizing the rock mass integrity grade is established. The ablation experiments show that architecture and class weight optimizations can increase the F₁ value of the SI-VGG19 model on the test set by 2.0% and 1.6%, respectively. The F₁ value of the proposed model on the test set is as high as 0.915, which is 13.1%, 8.3%, 3.9%, 10.2%, 9.0%, 19.7%, 10.9%, and 18.0% higher than that of AlexNet, Xception, VGG19, ResNet50, InceptionV3, MobileNetV2, DenseNet121, and EfficientNetV2B3 models, respectively. Therefore, the proposed method is excellent which can provide guidance for disaster warning and construction optimization of TBM tunnel.

Keywords

Tunnel boring machine (TBM) / Rock mass integrity / Rock fragment image / Convolutional neural network / Attention mechanism

Declaration of competing interest

Quansheng Liu is an Editorial Board Member for this journal, and was not involved in the editorial review or the decision to publish this article. The authors declare the following personal re-lationships which may be considered as potential competing inter-ests: Jiquan Zi and Xiya Li are currently employed by Sinohydro Bu-reau 14 Co., Ltd. Other authors declare that there are no competing interests.

CRediT authorship contribution statement

Honggan Yu: Writing -review & editing, Writing -original draft, Validation, Methodology, Conceptualization. Shuzhan Xu: Writing -re-view & editing, Visualization, Software, Data curation. Penghai Deng: Validation, Investigation, Formal analysis. Xin Yin: Writing -review & editing, Visualization, Validation, Project administration, Methodology. Jiquan Zi: Validation, Investigation, Formal analysis, Data curation. Xiya Li: Visualization, Software, Resources. Quansheng Liu: Writing -review & editing, Validation, Supervision.

Acknowledgements

This research is supported by the National Natural Science Foun-dation of China (Grant No. 42507232), Natural Science Foundation of Hubei Province (Grant No. 2025AFB137), and Fundamental Research Funds for the Central Universities (Grant No. 2042024rs0001).

References

[1] H. Yu, C. Qin, J. Tao, C. Liu, Q. Liu, A multi-channel decoupled deep neural net-work for tunnel boring machine torque and thrust prediction, Tunn. Undergr. Space Technol. 133 (2023) 104949, doi:10.1016/j.tust.2022.104949.

[2] H. Yu, Y. Bo, Q. Liu, X. Yang, S. Xu, X. Huang, Monthly advance rate estimation of hard rock tunnel boring machine based on rock mass classification and data aug-mentation, Undergr. Space 23 (2025) 175-192, doi:10.1016/j.undsp.2025.02.002.

[3] G. Zhao, Z. Li, Y. Chen, Q. Gong, S. Li, X. Wei, A digital twin-based multiscale frame-work for predicting full-scale TBM rock cutting performance from miniature point load tests, Smart Undergr. Eng. 1 (2025) 51-63, doi:10.1016/j.sue.2025.05.001.

[4] H. Yu, S. Xu, X. Yin, M. Hasanipanah, A sliding window feature extraction and weight adaptive random forest-based method for TBM tunnel rock mass grade identification, Expert Syst. Appl. 311 (2026) 131336, doi:10.1016/j.eswa.2026.131336.

[5] H. Yu, H. Sun, J. Tao, C. Qin, D. Xiao, Y. Jin, C. Liu, A multi-stage data augmentation and AD-ResNet-based method for EPB utilization factor prediction, Automat. Constr. 147 (2023) 104734, doi:10.1016/j.autcon.2022.104734.

[6] Z. Xu, T. Yu, S. Li, P. Lin, W. Ma, T. Han, S. Li, Intelligent identification of lithology and adverse geology: a state-of-the-art review, Smart Undergr. Eng. 1 (2025) 3-25, doi:10.1016/j.sue.2025.04.001.

[7] H. Wang, Z. Huang, C. Yang, T. Zhu, Application of TBM adaptability index in Gaoligong Mountain Tunnel, Water Resourc. Hydropower Eng. 52 (2021) 180-186, doi:10.13928/j.cnki.wrahe.2021.S1.033.

[8] J. Huang, Optimization and study on adaptability of tunneling parameter for open-type TBM in zone of high abradable hard rock, Water Resourc. Hydropower Eng. 48 (2017) 90-95, doi:10.13928/j.cnki.wrahe.2017.08.018.

[9] H. Yu, J. Tao, S. Huang, C. Qin, D. Xiao, C. Liu, A field parameters-based method for real-time wear estimation of disc cutter on TBM cutterhead, Automat. Constr. 21 (2021) 103603, doi:10.1016/j.autcon.2021.103603.

[10] Z. Li, Z. Zhou, Z. Tan, Z. Guo, L. Zhang, Development and experimental verification of a new model for predicting laser-induced rock cutting slit depth, Int. J. Rock Mech. Min. 177 (2024) 105722, doi:10.1016/j.ijrmms.2024.105722.

[11] Z. Zhou, Z. Li, Z. Tan, K. Lei, L. Zhang, Optimized decision-making for tunnel boring machine control parameters, Rock Mech. Rock Eng. 58 (2025) 5711-5731, doi:10.1007/s00603-025-04451-6.

[12] Z. Zhou, Z. Tan, Z. Li, D. Ma, L. Zhang, TBM cutter wear under high strength surrounding rock conditions: a case study from the second phase of the north-ern Xinjiang water supply project, Rock Mech. Rock Eng. 54 (2021) 5023-5039, doi:10.1007/s00603-021-02545-5.

[13] X. Zhang, W. Xie, K. Cai, Q. Liu, J. Wu, W. Li, Evaluation of rock muck using image analysis and its application in the TBM tunneling, Tunn. Undergr. Space Technol. 113 (2021) 103974, doi:10.1016/j.tust.2021.103974.

[14] A. Rispoli, A.M. Ferrero, M. Cardu, A. Farinetti, Determining the particle size of debris from a tunnel boring machine through photographic analysis and comparison between excavation performance and rock mass properties, Rock Mech. Rock Eng. 50 (2017) 2805-2816, doi:10.1007/s00603-017-1256-5.

[15] C. Yan, W. Zhao, T. Wang, Z. Gao, X. Xiao, J. Qin, Y. Jiang, Evaluation indicators and classification system for rock mass boreability based on rock mass chip image processing, Eng. Geol. 354 (2025) 108176, doi:10.1016/j.enggeo.2025.108176.

[16] E. Farrokh, J. Rostami, Correlation of tunnel convergence with TBM operational pa-rameters and chip size in the Ghomroud tunnel, Iran, Tunn. Undergr. Space Technol. 23 (2008) 700-710, doi:10.1016/j.tust.2008.01.005.

[17] Z. Dong, X. Li, G. Tian, Y. Pang, L. Wang, Classification prediction of surround-ing rock based on TBM muck images, J. Basic Sci. Eng. 31 (2023) 1540-1551, doi:10.16058/j.issn.1005-0930.2023.06.012.

[18] S. Heydari, J.K. Hamidi, M. Monjezi, A. Eftekhari, An investigation of the relation-ship between muck geometry, TBM performance, and operational parameters: a case study in Golab II water transfer tunnel, Tunn. Undergr. Space Technol. 88 (2019) 73-86, doi:10.1016/j.tust.2018.11.043.

[19] X. Zhou, Q. Gong, Y. Liu, L. Yin, Automatic segmentation of TBM muck images via a deep-learning approach to estimate the size and shape of rock chips, Automat. Constr. 126 (2021) 103685, doi:10.1016/j.autcon.2021.103685.

[20] K. Lei, Z. Tan, X. Wang, Z. Zhou, Semi-symmetrical, fully convolutional masked autoencoder for TBM muck image segmentation, Symmetry 16 (2024) 222, doi:10.3390/sym16020222.

[21] X. Li, H. Wang, H. Yang, Hyperspectral and image information of rock slag: an ef-fective means to evaluate TBM tunneling efficiency, Tunn. Undergr. Space Technol. 166 (2025) 107020, doi:10.1016/j.tust.2025.107020.

[22] L. Chen, Z. Liu, H. Su, F. Lin, W. Mao, Automated rock mass condition assessment during TBM tunnel excavation using deep learning, Sci. Rep.-Uk 12 (2022) 1722, doi:10.1038/s41598-022-05727-5.

[23] L. Chen, Q. Zheng, Z. Liu, W. Mao, F. Lin, Self-intersection attention pooling based classification for rock recognition, in: Proceedings of 16th International Conference on Control, Automation, Robotics and Vision, 2020, pp. 1210-1215, doi:10.1109/ICARCV50220.2020.9305305.

[24] G. Huang, C. Qin, H. Wang, C. Liu, TBM rock fragmentation classifica-tion using an adaptive spot denoising and contour-texture decomposition attention-based method, Tunn. Undergr. Space Technol. 161 (2025) 106498, doi:10.1016/j.tust.2025.106498.

[25] L. Huang, Q. Gong, J. Wang, H. Ma, X. Zhou, X. Xie, H. Song, A real-time rock mass class identification model of the tunnel face based on TBM tunneling and the corresponding muck characteristic parameters, Int. J. Rock Mech. Min. 188 (2025) 106057, doi:10.1016/j.ijrmms.2025.106057.

[26] W. Qiao, Y. Zhao, Y. Xu, Y. Lei, Y. Wang, S. Yu, H. Li, Deep learning-based pixel-level rock fragment recognition during tunnel excavation using in-stance segmentation model, Tunn. Undergr. Space Technol. 115 (2021) 104072, doi:10.1016/j.tust.2021.104072.

[27] W. Xie, X. Zhang, X. Liu, C. Xu, X. Li, D. Song, Q. Ma, N. Hu, Real-time perception of rock-machine interaction information in TBM tunnelling us-ing muck image analysis, Tunn. Undergr. Space Technol. 136 (2023) 105096, doi:10.1016/j.tust.2023.105096.

[28] Z. Xue, L. Chen, Z. Liu, F. Lin, W. Mao, Rock segmentation visual system for assisting driving in TBM construction, Mach. Vision Appl. 32 (2021) 77, doi:10.1007/s00138-021-01203-8.

[29] Q. Gong, X. Zhou, Y. Liu, B. Han, L. Yin, Development of a real-time muck analysis system for assistant intelligence TBM tunnelling, Tunn. Undergr. Space Technol. 107 (2021) 103655, doi:10.1016/j.tust.2020.103655.

[30] F. Li, Q. Liu, Y. Pan, Y. Bo, An attention-enhanced ResNet model for classifi-cation of TBM rock chips, Tunn. Undergr. Space Technol. 165 (2025) 106916, doi:10.1016/j.tust.2025.106916.

[31] Y. Liu, D. Wang, J. Hu, G. Zhu, Classifying rock fragments produced by tunnel boring machine using optimized convolutional neural network, Rock Mech. Rock Eng. 57 (2024) 1765-1780, doi:10.1007/s00603-023-03623-6.

[32] Z. Yang, B. He, Y. Liu, D. Wang, G. Zhu, Classification of rock fragments produced by tunnel boring machine using convolutional neural networks, Automat. Constr. 125 (2021) 103612, doi:10.1016/j.autcon.2021.103612.

[33] G. Huang, C. Qin, T. Zhong, C. Liu, A novel multi-scale hybrid con-nected neural network for anti-noise rock fragmentation classification of tun-nel boring machine, Tunn. Undergr. Space Technol. 161 (2025) 106555, doi:10.1016/j.tust.2025.106555.

[34] Ministry of Water Resources of the People’s Republic of China, GB/T 50218- 2014: Standard for Engineering Classification of Rock Masses. China Planning Press, Bei-jing, China. (in Chinese).

[35] J. Liu, Z. Zhou, Y. Pan, Q. Liu, Characterization of particle size distribution of rock chips and evaluation of rock breaking efficiency of TBM based on multi-fractal theory: a case study, Tunn. Undergr. Space Technol. 158 (2025) 106418, doi:10.1016/j.tust.2025.106418.

[36] A. Azizi, H. Moomivand, A new approach to represent impact of discontinuity spac-ing and rock mass description on the median fragment size of blasted rocks us-ing image analysis of rock mass, Rock Mech. Rock Eng. 54 (2021) 2013-2038, doi:10.1007/s00603-020-02360-4.

[37] Q. Guo, S. Yang, Y. Wang, H. Zhang, Size measurement of blasted rock fragments based on FRRSnet+, Measurement 218 (2023) 113207, doi:10.1016/j.measurement.2023.113207.

[38] J. Zhao, D. Li, Y. Yu, Identification of rock fragments after blasting by us-ing deep learning-based segment anything model, Minerals 14 (2024) 654, doi:10.3390/min14070654.

[39] Y. Jiang, Z. Pan, X. Zhang, Tunnel infrastructure health management: a state-of-the-art review on defect development mechanism and robot-aided inspec-tion system, Smart Undergr. Eng. 1 (2025) 26-39, doi:10.1016/j.sue.2025.05.003.

[40] D. Nimma, A. Uddagiri, Advancements in deep learning architectures for image recognition and semantic segmentation, Int. J. Adv. Comput. Sc. 15 (2024) 1172, doi:10.14569/ijacsa.2024.01508114.

[41] G. Cui, L. Zhang, Improved faster region convolutional neural network algorithm for UAV target detection in complex environment, Results Eng. 23 (2024) 102487, doi:10.1016/j.rineng.2024.102487.

[42] N. Manakitsa, G.S. Maraslidis, L. Moysis, G.F. Fragulis, A review of machine learn-ing and deep learning for object detection, semantic segmentation, and human action recognition in machine and robotic vision, Technologies 12 (2024) 15, doi:10.3390/technologies12020015.

[43] K. Simonyan, A. Zisserman, Very Deep Convolutional Networks for Large-Scale Image Recognition, 2014, doi:10.48550/arXiv.1409.1556.arXivpreprintarXiv:1409.1556.

[44] Z. Li, B. Li, S.G. Jahng, C. Jung, Improved VGG algorithm for visual prosthe-sis image recognition, IEEE Access 12 (2024) 45727-45739, doi:10.1109/AC-CESS.2024.3380839.

[45] C. He, H. Sadeghpour, Y. Shi, B. Mishra, S. Roshankhah, Mapping dis-tribution of fractures and minerals in rock samples using Res-VGG-UNet and threshold segmentation methods, Comput. Geotech. 175 (2024) 106675, doi:10.1016/j.compgeo.2024.106675.

[46] P.K. Sahoo, M.K. Panda, U. Panigrahi, G. Panda, P. Jain, M.S. Islam, An improved VGG-19 network induced enhanced feature pooling for precise moving object detec-tion in complex video scenes, IEEE Access 12 (2024) 45847-45864, doi:10.1109/AC-CESS.2024.3381612.

[47] C. Szegedy, W. Liu, Y. Jia, P. Sermanet, S. Reed, D. Anguelov, D. Erhan, V. Van- houcke, A. Rabinovich, Going deeper with convolutions, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2015, pp. 1-9, doi:10.48550/arXiv.1409.4842.

[48] S. Ioffe, C. Szegedy, Batch normalization: accelerating deep network training by reducing internal covariate shift,in: International Conference on Machine Learning, 2015, pp. 448-456. http://proceedings.mlr.press/v37/ioffe15.pdf.

[49] C. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens, Z. Wojna, Rethinking the in-ception architecture for computer vision, in: Proceedings of the IEEE Con-ference on Computer Vision and Pattern Recognition, 2016, pp. 2818-2826. https://www.cv-foundation.org/openaccess/content_cvpr_2016/papers/Szegedy_Rethinking_the_Inception_CVPR_2016_paper.pdf.

[50] M. Guo, T. Xu, J. Liu, Z. Liu, P. Jiang, T. Mu, S. Zhang, R.R. Martin, M. Cheng, S. Hu, Attention mechanisms in computer vision: a survey, Comput. Vis. Media 8 (2022) 331-368, doi:10.1007/s41095-022-0271-y.

[51] G. Brauwers, F. Frasincar, A general survey on attention mechanisms in deep learning, IEEE T. Knowl. Data En. 35 (2021) 3279-3298, doi:10.1109/TKDE.2021.3126456.

[52] J. Hu, L. Shen, G. Sun, Squeeze-and-excitation networks, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2018, pp. 7132-7141, doi:10.48550/arXiv.1709.01507.

[53] A. Krizhevsky, I. Sutskever, G.E. Hinton, ImageNet classification with deep convo-lutional neural networks, Commun. ACM 60 (2017) 84-90, doi:10.1145/3065386.

[54] F. Chollet, Xception: deep learning with depthwise separable convolutions, in: Pro-ceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017, pp. 1251-1258, doi:10.48550/arXiv.1610.02357.

[55] K. He, X. Zhang, S. Ren, J. Sun,Deep residual learning for image recogni-tion, in:Proceedings of the IEEE Conference on Computer Vision and Pat-tern Recognition, 2016, pp. 770-778. https://openaccess.thecvf.com/content_cvpr_2016/papers/He_Deep_Residual_Learning_CVPR_2016_paper.pdf.

[56] C. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens, Z. Wojna,Rethinking the inception architecture for computer vision, in:Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016, pp. 2818-2826. https://www.cv-foundation.org/openaccess/content_cvpr_2016/papers/Szegedy_Rethinking_the_Inception_CVPR_2016_paper.pdf.

[57] M. Sandler, A. Howard, M. Zhu, A. Zhmoginov, L. Chen,Mobilenetv2: in-verted residuals and linear bottlenecks, in: Proceedings of the IEEE Confer-ence on Computer Vision and Pattern Recognition, 2018, pp. 4510-4520, doi:10.48550/arXiv.1801.04381.

[58] G. Huang, Z. Liu, L.V.D. Maaten, K.Q. Weinberger, Densely connected convolutional networks, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017, pp. 4700-4708, doi:10.48550/arXiv.1608.06993.

[59] M. Tan, Q. Le, Efficientnet: rethinking model scaling for convolutional neural net-works,in: Proceedings of the International Conference on Machine Learning, 97, 2019, pp. 6105-6114. http://proceedings.mlr.press/v97/tan19a/tan19a.pdf.