Deep learning of rock microscopic images for intelligent lithology identification: Neural network comparison and selection
An intelligent lithology identification method is proposed based on deep learning of the rock microscopic images. Based on the characteristics of rock images in the dataset, we used Xception, MobileNet_v2, Inception_ResNet_v2, Inception_v3, Densenet121, ResNet101_v2, and ResNet-101 to develop micros...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2022-08-01
|
| Series: | Journal of Rock Mechanics and Geotechnical Engineering |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1674775522001202 |
| _version_ | 1811315922891177984 |
|---|---|
| author | Zhenhao Xu Wen Ma Peng Lin Yilei Hua |
| author_facet | Zhenhao Xu Wen Ma Peng Lin Yilei Hua |
| author_sort | Zhenhao Xu |
| collection | DOAJ |
| description | An intelligent lithology identification method is proposed based on deep learning of the rock microscopic images. Based on the characteristics of rock images in the dataset, we used Xception, MobileNet_v2, Inception_ResNet_v2, Inception_v3, Densenet121, ResNet101_v2, and ResNet-101 to develop microscopic image classification models, and then the network structures of seven different convolutional neural networks (CNNs) were compared. It shows that the multi-layer representation of rock features can be represented through convolution structures, thus better feature robustness can be achieved. For the loss function, cross-entropy is used to back propagate the weight parameters layer by layer, and the accuracy of the network is improved by frequent iterative training. We expanded a self-built dataset by using transfer learning and data augmentation. Next, accuracy (acc) and frames per second (fps) were used as the evaluation indexes to assess the accuracy and speed of model identification. The results show that the Xception-based model has the optimum performance, with an accuracy of 97.66% in the training dataset and 98.65% in the testing dataset. Furthermore, the fps of the model is 50.76, and the model is feasible to deploy under different hardware conditions and meets the requirements of rapid lithology identification. This proposed method is proved to be robust and versatile in generalization performance, and it is suitable for both geologists and engineers to identify lithology quickly. |
| first_indexed | 2024-04-13T11:38:50Z |
| format | Article |
| id | doaj.art-bcd6632a73414a4faef8b2ba89caea3c |
| institution | Directory Open Access Journal |
| issn | 1674-7755 |
| language | English |
| last_indexed | 2024-04-13T11:38:50Z |
| publishDate | 2022-08-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Journal of Rock Mechanics and Geotechnical Engineering |
| spelling | doaj.art-bcd6632a73414a4faef8b2ba89caea3c2022-12-22T02:48:21ZengElsevierJournal of Rock Mechanics and Geotechnical Engineering1674-77552022-08-0114411401152Deep learning of rock microscopic images for intelligent lithology identification: Neural network comparison and selectionZhenhao Xu0Wen Ma1Peng Lin2Yilei Hua3Geotechnical and Structural Engineering Research Center, Shandong University, Jinan, 250061, China; School of Qilu Transportation, Shandong University, Jinan, 250061, China; Corresponding author. Geotechnical and Structural Engineering Research Center, Shandong University, Jinan, 250061, China.Geotechnical and Structural Engineering Research Center, Shandong University, Jinan, 250061, ChinaGeotechnical and Structural Engineering Research Center, Shandong University, Jinan, 250061, ChinaInstitute of Robotics and Intelligent Systems, Wuhan University of Science and Technology, Wuhan, 430000, ChinaAn intelligent lithology identification method is proposed based on deep learning of the rock microscopic images. Based on the characteristics of rock images in the dataset, we used Xception, MobileNet_v2, Inception_ResNet_v2, Inception_v3, Densenet121, ResNet101_v2, and ResNet-101 to develop microscopic image classification models, and then the network structures of seven different convolutional neural networks (CNNs) were compared. It shows that the multi-layer representation of rock features can be represented through convolution structures, thus better feature robustness can be achieved. For the loss function, cross-entropy is used to back propagate the weight parameters layer by layer, and the accuracy of the network is improved by frequent iterative training. We expanded a self-built dataset by using transfer learning and data augmentation. Next, accuracy (acc) and frames per second (fps) were used as the evaluation indexes to assess the accuracy and speed of model identification. The results show that the Xception-based model has the optimum performance, with an accuracy of 97.66% in the training dataset and 98.65% in the testing dataset. Furthermore, the fps of the model is 50.76, and the model is feasible to deploy under different hardware conditions and meets the requirements of rapid lithology identification. This proposed method is proved to be robust and versatile in generalization performance, and it is suitable for both geologists and engineers to identify lithology quickly.http://www.sciencedirect.com/science/article/pii/S1674775522001202Deep learningRock microscopic imagesAutomatic classificationLithology identification |
| spellingShingle | Zhenhao Xu Wen Ma Peng Lin Yilei Hua Deep learning of rock microscopic images for intelligent lithology identification: Neural network comparison and selection Journal of Rock Mechanics and Geotechnical Engineering Deep learning Rock microscopic images Automatic classification Lithology identification |
| title | Deep learning of rock microscopic images for intelligent lithology identification: Neural network comparison and selection |
| title_full | Deep learning of rock microscopic images for intelligent lithology identification: Neural network comparison and selection |
| title_fullStr | Deep learning of rock microscopic images for intelligent lithology identification: Neural network comparison and selection |
| title_full_unstemmed | Deep learning of rock microscopic images for intelligent lithology identification: Neural network comparison and selection |
| title_short | Deep learning of rock microscopic images for intelligent lithology identification: Neural network comparison and selection |
| title_sort | deep learning of rock microscopic images for intelligent lithology identification neural network comparison and selection |
| topic | Deep learning Rock microscopic images Automatic classification Lithology identification |
| url | http://www.sciencedirect.com/science/article/pii/S1674775522001202 |
| work_keys_str_mv | AT zhenhaoxu deeplearningofrockmicroscopicimagesforintelligentlithologyidentificationneuralnetworkcomparisonandselection AT wenma deeplearningofrockmicroscopicimagesforintelligentlithologyidentificationneuralnetworkcomparisonandselection AT penglin deeplearningofrockmicroscopicimagesforintelligentlithologyidentificationneuralnetworkcomparisonandselection AT yileihua deeplearningofrockmicroscopicimagesforintelligentlithologyidentificationneuralnetworkcomparisonandselection |