WCSE 2019 SUMMER ISBN: 978-981-14-1684-2
DOI: 10.18178/wcse.2019.06.052

Identifying Rock Thin Section Based on Convolutional Neural Networks

Ren Wei, Zhang Minghua, Zhang Sheng, Qiao Jihua, Huang Jinming

Abstract— This paper uses the convolutional neural networks to train and identify rock thin sections. There are three innovations in this method. Firstly the convolutional neural network is used in the field of geological experiment testing and it can automatically identify and classify the rock thin sections. Secondly the original image of rock thin sections are sliced and segmented, so that the convolutional neural network could learn more details of the rock mineral texture without damaging the original resolution of the image. Thirdly using other image enhancement techniques, such as random flipping and standardization, can expand the sample data set and enhance the robustness of the model. Finally the training model achieves the desired results.

Index Terms— deep learning, convolutional neural network, rock thin section, lithology identification

Ren Wei, Zhang Minghua, Zhang Sheng, Qiao Jihua, Huang Jinming
China Geological Survey Development and Research Center, CHINA

[Download]


Cite: Ren Wei, Zhang Minghua, Zhang Sheng, Qiao Jihua, Huang Jinming, "Identifying Rock Thin Section Based on Convolutional Neural Networks," Proceedings of 2019 the 9th International Workshop on Computer Science and Engineering, pp. 345-351, Hong Kong, 15-17 June, 2019.