Abstract
Rock classification plays an important role in rock mechanics, petrology, mining engineering, magmatic processes, and numerous other fields pertaining to geosciences. This study proposes a concatenated convolutional neural network (Con-CNN) method for classifying geologic rock types based on petrographic thin sections. Plane polarized light (PPL) and crossed polarized light (XPL) were used to acquire thin section images as the fundamental data. After conducting the necessary pre-processing, the PPL and XPL images as well as their comprehensive image developed by principal component analysis were sliced into small patches and were put into three CNNs, comprising the same structure for achieving a preliminary classification. Subsequently, these patches classification results of the CNNs were concatenated by using the maximum likelihood method to obtain a comprehensive classification result. Finally, a statistical revision was applied to fix the misclassification due to the proportion differences of minerals that were similar in appearance. In this study, there were 92 rock samples of 13 types giving 106 petrographic thin sections and 2208 petrographic thin section images, and finally 238,464 sliced image patches were used for the training and validation of the Con-CNN method. The 5-folds cross validation showed that the proposed method provides an overall accuracy of 89.97% and a kappa coefficient of 0.86, which facilitates the automation of rock classification in petrographic thin section images.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Aligholi S, Lashkaripour GR, Khajavi R, Razmara M (2017) Automatic mineral identification using color tracking. Pattern Recogn 65:164–174. https://doi.org/10.1016/j.patcog.2016.12.012
Asmussen P, Conrad O, Günther A, Kirsch M, Riller U (2015) Semi-automatic segmentation of petrographic thin section images using a “seeded-region growing algorithm” with an application to characterize weathered subarkose sandstone. Comput Geosci 83:89–99. https://doi.org/10.1016/j.cageo.2015.05.001
Baykan NA, Yılmaz N (2010) Mineral identification using color spaces and artificial neural networks. Comput Geosci 36:91–97. https://doi.org/10.1016/j.cageo.2009.04.009
Berrezueta E, Domínguez-Cuesta MJ, Rodríguez-Rey A (2019) Semi-automated procedure of digitalization and study of rock thin section porosity applying optical image analysis tools. Comput Geosci 124:14–26. https://doi.org/10.1016/j.cageo.2018.12.009
Borazjani O, Ghiasi-Freez J, Hatampour A (2016) Two intelligent pattern recognition models for automatic identification of textural and pore space characteristics of the carbonate reservoir rocks using thin section images. J Nat Gas Sci Eng 35:944–955. https://doi.org/10.1016/j.jngse.2016.09.048
Chatterjee S (2013) Vision-based rock-type classification of limestone using multi-class support vector machine. Appl Intell 39:14–27. https://doi.org/10.1007/s10489-012-0391-7
Cheng G, Yue Q, Qiang X (2018) Research on feasibility of convolution neural networks for rock thin sections image retrieval. In: 2018 2nd IEEE Advanced Information Management, Communicates, Electronic and Automation Control Conference, IMCEC 2018, Institute of Electrical and Electronics Engineers Inc., 2539–2542
Dong S, Zeng L, Xu C, Dowd P, Gao Z, Mao Z, Wang A (2019) A novel method for extracting information on pores from cast thin-section images. Comput Geosci 130:69–83. https://doi.org/10.1016/j.cageo.2019.05.003
Fauzi U (2011) An estimation of rock permeability and its anisotropy from thin sections using a renormalization group approach. Energ Source Part A 33:539–548. https://doi.org/10.1080/15567030903097038
Ghiasi-Freez J, Soleimanpour I, Kadkhodaie-Ilkhchi A, Ziaii M, Sedighi M, Hatampour A (2012) Semi-automated porosity identification from thin section images using image analysis and intelligent discriminant classifiers. Comput Geosci 45:36–45. https://doi.org/10.1016/j.cageo.2012.03.006
Hassanpour A, Kananian A, Barghi MA (2009) Minerals boundary detection in petrographic thin sections image using ArcGIS software. Iran J Crystallogr Miner 17:133–148
He K, Zhang X, Ren S, Sun J (2015) Deep residual learning for image recognition. arXiv:1512.03385. https://doi.org/10.1109/CVPR.2016.90
Izadi H, Sadri J, Mehran N (2013) A new approach to apply texture features in minerals identification in petrographic thin sections using ANNs. In: 2013 8th Iranian Conference on Machine Vision and Image Processing, (MVIP), Zanjan, 257–261. https://doi.org/10.1109/IranianMVIP.2013.6779990
Izadi H, Sadri J, Bayati M (2017) An intelligent system for mineral identification in thin sections based on a cascade approach. Comput Geosci 99:37–49. https://doi.org/10.1016/j.cageo.2016.10.010
Joseph S, Ujir H, Hipiny I (2017) Unsupervised classification of intrusive igneous rock thin section images using edge detection and colour analysis. In: 2017 IEEE International Conference on Signal and Image Processing Applications (ICSIPA). Institute of Electrical and Electronics Engineers Inc., 530–534. https://doi.org/10.1109/ICSIPA.2017.8120669
Krig S (2014) Image pre-processing. In: Krig S (ed) Computer vision metrics. Apress, Berkeley, pp 39–83
Ładniak M, Młynarczuk M (2015) Search of visually similar microscopic rock images. Comput Geosci 19:127–136. https://doi.org/10.1007/s10596-014-9459-2
Lecun Y, Bottou L, Bengio Y, Haffner PY (1998) Gradient-based learning applied to document recognition. Proc IEEE 86:2278–2324. https://doi.org/10.1109/5.726791
LeCun Y, Bengio Y, Hinton G (2015) Deep learning. Nature 521:436–444. https://doi.org/10.1038/nature14539
Li Y, Onasch CM, Guo Y (2008) GIS-based detection of grain boundaries. J Struct Geol 30:431–443. https://doi.org/10.1016/j.jsg.2007.12.007
Li N, Hao H, Gu Q, Wang D, Hu X (2017) A transfer learning method for automatic identification of sandstone microscopic images. Comput Geosci 103:111–121. https://doi.org/10.1016/j.cageo.2017.03.007
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. Zhongnan Daxue Xuebao (Ziran Kexue ban)/J Central South University (Science and Technology) 47:2375–2382
Marmo R, Amodio S, Tagliaferri R, Ferreri V, Longo G (2005) Textural identification of carbonate rocks by image processing and neural network: methodology proposal and examples. Comput Geosci 31:649–659. https://doi.org/10.1016/j.cageo.2004.11.016
Marques VG, Da Silva LRD, Carvalho BM, de Lucena LR, Vieira MM (2019) Deep learning-based pore segmentation of thin rock sections for aquifer characterization using color space reduction. 26th International Conference on Systems, Signals and Image Processing, IWSSIP 2019, IEEE Computer Society
Mingireanov Filho I, Vallin Spina T, Xavier Falcão A, Campane Vidal A (2013) Segmentation of sandstone thin section images with separation of touching grains using optimum path forest operators. Comput Geosci 57:146–157. https://doi.org/10.1016/j.cageo.2013.04.011
Młynarczuk M, Górszczyk A, Ślipek B (2013) The application of pattern recognition in the automatic classification of microscopic rock images. Comput Geosci 60:126–133. https://doi.org/10.1016/j.cageo.2013.07.015
Paszke A, Gross S, Massa F, Lerer A, Bradbury J, Chanan G, Killeen T, Lin Z, Gimelshein N, Antiga L, Desmaison A (2019) PyTorch: an imperative style, high-performance deep learning library. arXiv:1912.01703
Peng S, Hassan A, Loucks RG (2016) Permeability estimation based on thin-section image analysis and 2D flow modeling in grain-dominated carbonates. Mar Petrol Geol 77:763–775. https://doi.org/10.1016/j.marpetgeo.2016.07.024
Rabbani A, Assadi A, Kharrat R, Dashti N, Ayatollahi S (2017) Estimation of carbonates permeability using pore network parameters extracted from thin section images and comparison with experimental data. J Nat Gas Sci Eng 42:85–98. https://doi.org/10.1016/j.jngse.2017.02.045
Reedy CL (2006) Review of digital image analysis of petrographic thin sections in conservation research. J Am Inst Conserv 45:127–146. https://doi.org/10.1179/019713606806112531
Rubo RA, de Carvalho CC, Fontana Michelon M, dos Santos GR (2019) Digital petrography: mineralogy and porosity identification using machine learning algorithms in petrographic thin section images. J Pet Sci Eng 183:106382. https://doi.org/10.1016/j.petrol.2019.106382
Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale image recognition. Comp Sci. arXiv:1409.1556
Singh N, Singh TN, Tiwary A, Sarkar KM (2010) Textural identification of basaltic rock mass using image processing and neural network. Comput Geosci 14:301–310. https://doi.org/10.1007/s10596-009-9154-x
van den Berg EH, Meesters AGCA, Kenter JAM, Schlager W (2002) Automated separation of touching grains in digital images of thin sections. Comput Geosci 28:179–190. https://doi.org/10.1016/S0098-3004(01)00038-3
Wen Y, Zuo G, et al. (2019) A minerals boundary enhancement method in petrographic thin sections polarization images. 9th IEEE International Conference on Electronics Information and Emergency Communication, ICEIEC 2019, Institute of Electrical and Electronics Engineers Inc.
Xu S, Zhou Y (2018) Artificial intelligence identification of ore minerals under microscope based on deep learning algorithm. Acta Petrol Sin 34:3244–3252
Yesiloglu-Gultekin N, Keceli AS, Sezer EA, Can AB, Gokceoglu C, Bayhan H (2012) A computer program (TSecSoft) to determine mineral percentages using photographs obtained from thin sections. Comput Geosci 46:310–316. https://doi.org/10.1016/j.cageo.2012.01.001
Acknowledgements
This research was supported by the National Key R&D Program of China (NO. 2018YFB0505002), COMRA Major Project (NO. DY135-S1-01-01-03), and Zhejiang Provincial Natural Science Foundation of China (LY17D010006). The authors thank Prof. Zhongyue Shen, School of Earth Sciences, Zhejiang University, for his advice and assistance in processing the petrographic thin sections, and also thank the anonymous reviewers for their comments and suggestions that helped in enhancing the quality of our manuscript.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Su, C., Xu, Sj., Zhu, Ky. et al. Rock classification in petrographic thin section images based on concatenated convolutional neural networks. Earth Sci Inform 13, 1477–1484 (2020). https://doi.org/10.1007/s12145-020-00505-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12145-020-00505-1