Skip to main content
SpringerLink
Log in

Influence of Neural Network Structure on Rock Intelligent Classification Based on Structural and Tectonic Features of Rocks

  • Original Paper
  • Published:
Rock Mechanics and Rock Engineering Aims and scope Submit manuscript

Abstract

Fast and accurate intelligent classification of rocks is the most important prerequisite to ensure the unmanned construction of geological projects. In this paper, a dataset containing 660 images is established, classified based on the structure and the fabric characteristics of rocks, and investigate the recognition accuracy of different structural and tectonic features. Moreover, the accuracy, training time, and robustness of different neural network architectures on a multi-scale image set are compared by transfer learning. The results show that InceptionResnetV2 has the best robustness and high accuracy, which can reach 74.24%, 74.24%, and 72.73% on the three image sets respectively; and it is found that the Inception module significantly improves the recognition accuracy and the Fire module reduces the learning time by a factor of 5–10. Finally, MobileNetV2 was tested and it obtained the highest accuracy of 81.82%. In addition, MobileNetV2 has a strong ability to recognize typical rock structures such as bedding, columnar joints, and leaf veins, and can achieve a recognition accuracy of 95.83%.

Highlights

  • A data set based on the rock structure and fabric characteristics is built as the basis for lithology identification.

  • Based on transfer learning, structural and tectonic features of rocks can be recognized with an accuracy of 95.83%.

  • InceptionResnetV2 achieved the highest accuracy for recognition on a mixed-scale dataset, at 74.24%.

  • MobileNetV2 has powerful capabilities in identifying lithological structural features.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14
Fig. 15
Fig. 16
Fig. 17

Similar content being viewed by others

Abbreviations

e :

Napier constant

\(I\) :

Input layer matrix

W :

Neural network weight matrix

o :

Size of the image after the pooling layer

\(x_{i}\) :

Input value corresponding to the i-th neural network neurons

K :

Pooling kernel size

S :

String of the pooling operation

P :

Padding of the pooling operation

C :

Total number of input vectors entering SoftMax layer

References

  • Amosu A, Sun Y (2017) WheelerLab: an interactive program for sequence stratigraphic analysis of seismic sections, outcrops and well sections and the generation of chronostratigraphic sections and dynamic chronostratigraphic sections. SoftwareX 6:19–24

    Article  Google Scholar 

  • Bahmani AA, Riahi MA, Ramin N (2020) Detection of stratigraphic traps in the Asmari Formation using seismic attributes, petrophysical logs, and geological data in an oil field in the Zagros basin, Iran. J Petrol Sci Eng 194:107517

    Article  Google Scholar 

  • Baker BR, Gessner K, Holden E-J, Squelch AP (2008) Automatic detection of anisotropic features on rock surfaces. Geosphere 4(2):418–428

    Article  Google Scholar 

  • Castafio R, Anderson, R, Estlin, T, DeCoste, D, Fisher, F, Gaines, D, Mazzoni, D, Judd, M (Year) Rover traverse science for increased mission science return. In: 2003 IEEE aerospace conference. 8–15 March Big Sky, MT, United States, pp 3629–3636

  • Deepnarain N, Nasr M, Kumari S, Stenström TA, Reddy P, Pillay K, Bux F (2020) Artificial intelligence and multivariate statistics for comprehensive assessment of filamentous bacteria in wastewater treatment plants experiencing sludge bulking. Environ Technol Innov 19:100853

    Article  Google Scholar 

  • Dramsch JS, Lüthje M (2018) Deep-learning seismic facies on state-of-the-art CNN architectures. In: Seg technical program expanded abstracts 2018. Society of Exploration Geophysicists, pp 2036–2040

  • Duan Y, Zheng X, Hu L, Sun L (2019) Seismic facies analysis based on deep convolutional embedded clustering. Geophysics 84(6):IM87–IM97

  • Dunlop H (2006) Automatic rock detection and classification in natural scenes. Masters Thesis, Carnegie Mellon University.

  • Fan L, Zhao W, Gong ZN (2012) Correspondence analysis of rock spectra based on continuum removing. J Jilin Univ 42(2):575–582

    Google Scholar 

  • Feng YX, Gong X, Xu YY, Xie Z, Cai HH, Lv X (2019) Lithology recognition based on fresh rock images and twins convolution neural network. Geogr Geo-Inf Sci 5:89–94 ((in Chinese))

    Google Scholar 

  • Galdames FJ, Perez CA, Estévez PA, Adams M (2019) Rock lithological classification by hyperspectral, range 3D and color images. Chemom Intell Lab Syst 189:138–148

    Article  Google Scholar 

  • Gotlieb CC, Kreyszig H (1990) Texture descriptors based on co-occurrence matrices. Comput vis Gr Image Process 51(1):70–86

    Article  Google Scholar 

  • Guo Q, Zhou Y, Cao S, Qiu Z, Xu Z, Zhang Y (2010) Study on mineralogy of Guangning jade. Acta Sci Natur Univ Sunyatseni 49(3):146–151

    Google Scholar 

  • Han S, Wang S, Tang Z, Tan K, Duan X, He H, Feng Z, Xie Y (2020) Integrated geophysical exploration of the coupling of a concealed rock body and metallogenic structures—Ag–Pb–Zn mining area case study in Jilinbaolige, Inner Mongolia, China. J Appl Geophys 178:104048

    Article  Google Scholar 

  • Iwaszenko S, Smoliński A (2021) Texture features for bulk rock material grain boundary segmentation. J King Saud Univ Eng Sci 33(2):95–103

    Google Scholar 

  • Jing LJ, Li JB, Zhang N, Chen S, Yang C, Cao HB (2021) A TBM advance rate prediction method considering the effects of operating factors. Tunn Undergr Sp Technol 107:103620

    Article  Google Scholar 

  • Julesz B (1962) Visual pattern discrimination. IRE Trans Inf Theory 8(2):84–92

    Article  Google Scholar 

  • Julesz B (1981) Textons, the elements of texture perception, and their interactions. Nature 290(5802):91–97

    Article  Google Scholar 

  • Kereszturi A (2011) Geologic field work on Mars: distance and time issues during surface exploration. Acta Astronaut 68(11):1686–1701

    Article  Google Scholar 

  • Kim T, Hong K, Byun H (2021) The feature generator of hard negative samples for fine-grained image recognition. Neurocomputing 439:374–382

    Article  Google Scholar 

  • LeCun Y, Boser B, Denker JS, Henderson D, Howard RE, Hubbard W, Jackel LD (1989) Backpropagation applied to handwritten zip code recognition. Neural Comput 1(4):541–551

    Article  Google Scholar 

  • Lepisto L, Kunttu I, Autio J, Visa A (2003) Classification method for colored natural textures using gabor filtering. In: 12th International conference on image analysis and processing, 2003. Proceedings, pp 397–401

  • Leung T, Malik J (2001) Representing and recognizing the visual appearance of materials using three-dimensional textons. Int J Comput vis 43(1):29–44

    Article  Google Scholar 

  • Li MC, Fu JK, Zhang Y, Liu C (2020a) Intelligent recognition and analysis method of rock lithology classification based on coupled rock images and hammering audios. Chin J Rock Mech Eng 39(5):996–1004 ((in Chinese))

    Google Scholar 

  • Li N, Hao H, Jiang Z, Jiang F, Guo R, Gu Q, Hu X (2020b) A multi-task multi-class learning method for automatic identification of heavy minerals from river sand. Comput Geosci 135:104403

    Article  Google Scholar 

  • Liu Y, Cheng G, Ma W, Guo C (2016) Rock classification based on features form color space and morphological gradient of rock thin section image. J Cent South Univ (science and Technology) 47:2375–2382

    Google Scholar 

  • Purswani P, Karpyn ZT, Enab K, Xue Y, Huang X (2020) Evaluation of image segmentation techniques for image-based rock property estimation. J Petrol Sci Eng 195:107890

    Article  Google Scholar 

  • Puzyrev V, Elders C (2020) Unsupervised seismic facies classification using deep convolutional autoencoder. arXiv preprint, arXiv:.01995. https://arxiv.org/abs/2008.01995. Accessed 1 Oct 2021

  • Sehsah O, Ghazzawi A, Vie G, Al-Tajar T, Ali A, Al-Mohammed A, Itani M, Ullah S, Escalera H, Balka M (2017) Intelligent drilling system: expanding the envelope of wired drill pipe. In: Abu Dhabi international petroleum exhibition & conference

  • Tchokpon KG, Kaki C, Kourouma M, Yalo N (2020) Detection of gold-bearing quartz veins in the meta-sedimentary formation in the North-Eastern Guinea using remote sensing and geophysical exploration. J Afr Earth Sci 168:103869

    Article  Google Scholar 

  • Thompson D, Niekum S, Smith T, Wettergreen D (2005a) Automatic detection and classification of features of geologic interest. In: 2005 IEEE aerospace conference, pp 366–377

  • Thompson DR, Smith T, Wettergreen D (2005b) Data mining during rover traverse: from images to geologic signatures. In: 8th International symposium on artificial intelligence, robotics and automation in space, pp 31–32

  • Tuceryan M, Jain AK (1998) Texture analysis. In: Handbook of pattern recognition and computer vision, 2nd edn. World Scientific Publishing, Singapore, pp 235–276. https://doi.org/10.1142/9789814343138_0010

  • Valentín MB, Bom CR, Coelho JM, Correia MD, Márcio P, Marcelo P, Faria EL (2019) A deep residual convolutional neural network for automatic lithological facies identification in Brazilian pre-salt oilfield wellbore image logs. J Petrol Sci Eng 179:474–503

    Article  Google Scholar 

  • Varma M, Zisserman A (2005) A statistical approach to texture classification from single images. Int J Comput vis 62(1–2):61–81

    Article  Google Scholar 

  • Wang C, Li Y, Fan G, Chen F, Wang W (2018a) Quick recognition of rock images for mobile applications. J Eng Sci Technol Rev 11(4):111–117

    Article  Google Scholar 

  • Wang Z, Liu X, Fu Y, Yuan W, Miao L (2018b) Comprehensive screening method of “ultrasonic-rebound-density” for two kinds of rock specimens. Chin J Rock Mech Eng 37:3575–3583

    Google Scholar 

  • Xia Y, Li Q, Deng C, Yao J, Long B (2021) Rapid identification method for lithology of tunnel based on lightweight model. J Southw Jiaotong Univ 56(2):420–427

    Google Scholar 

  • Xu G, Zhu X, Fu D, Dong J, Xiao X (2017) Automatic land cover classification of geo-tagged field photos by deep learning. Environ Model Softw 91:127–134

    Article  Google Scholar 

  • Xu Z, Shi H, Lin P, Liu T (2021) Integrated lithology identification based on images and elemental data from rocks. J Petrol Sci Eng 205:108853

    Article  Google Scholar 

  • Yan D, Xu H, Xu Z, Lei Z, Tian M, Cheng L, Ma Y, Wang Z, Ostadhassan M (2020) Sedimentary architecture of hyperpycnal flow deposits: cretaceous Sangyuan outcrop, from the Luanping Basin, North East China. Mar Pet Geol 121:104593

    Article  Google Scholar 

  • Zhang Y, Wang G, Li M, Han S (2018) Automated classification analysis of geological structures based on images data and deep learning model. Appl Sci 8(12):2493

    Article  Google Scholar 

  • Zhao T (2018) Seismic facies classification using different deep convolutional neural networks. In: SEG Technical program expanded abstracts 2018. Society of Exploration Geophysicists, pp 2046–2050

Download references

Acknowledgements

This work was financially supported by the National Key Research and Development Plan (Grant no. 2018YFC1504902), the National Natural Science Foundation (Grant nos. 52079068, 41941019), and the State Key Laboratory of Hydroscience and Hydraulic Engineering (Grant no. 2021-KY-04).

Author information

Authors and Affiliations

  1. State Key Laboratory of Hydro-Science and Engineering, Tsinghua University, Beijing, 100084, China

    Mingyang Wang, Enzhi Wang, Xiaoli Liu, Zheng Wang & Congcong Wang

Authors
  1. Mingyang Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Enzhi Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Xiaoli Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Zheng Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Congcong Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Xiaoli Liu.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Wang, M., Wang, E., Liu, X. et al. Influence of Neural Network Structure on Rock Intelligent Classification Based on Structural and Tectonic Features of Rocks. Rock Mech Rock Eng 55, 5415–5432 (2022). https://doi.org/10.1007/s00603-022-02907-7

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00603-022-02907-7

Keywords

Search

Navigation