Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Fig 1.

Flowchart of the overall procedure of the proposed method.

More »

Fig 1 Expand

Fig 2.

Demonstrate the automated processing of core images.

Drill core images and lithology logs obtained from the geological drill holes were first uploaded to the geological database. The core images were gained from the database, and the cropping program made a judgment on the image and cropped it by center cropping. In total, 10 lithology categories were obtained, namely: (a) Diabase; (b) Diorite; (c) Gneiss; (d) Granite; (e) Limestone; (f) Marble; (g) Monzonite; (h) Mudstone; (i) Shale; and (j) Siltstone.

More »

Fig 2 Expand

Fig 3.

Example of non-core sections images.

The parts were easy to cause misclassification by the CNN model, such as red marks and crushing structures.

More »

Fig 3 Expand

Fig 4.

Ten typical samples in each label.

The non-core itself was discarded and the core images were annotated according to the geologist’s lithology logs.

More »

Fig 4 Expand

Fig 5.

The structure of ResNeSt block.

c,h,w are the number of channels, height, and width of the input featuremap, respectively. Cardinal k is the k-th cardinal group and Split r is the r-th split.

More »

Fig 5 Expand

Fig 6.

Diagram of learning rate decay.

The maximum learning rate is 4e−5, the minimum learning rate is 1e−7, and the total number of epochs is 100.

More »

Fig 6 Expand

Fig 7.

Fine-tuning.

Ffc represents full connection. The weight parameters of the output layer were initialized by Xavier random initialization.

More »

Fig 7 Expand

Fig 8.

The effect of whether to use data augmentation on model training.

Data augmentation can effectively alleviate overfitting and improve the generalization of the model.

More »

Fig 8 Expand

Table 1.

Compare the experimental results of the model at three batch sizes.

More »

Table 1 Expand

Fig 9.

Total loss of four different CNN architectures during training (a) and validation (b).

More »

Fig 9 Expand

Fig 10.

Accuracy of four different CNN architectures during training (a) and validation (b).

More »

Fig 10 Expand

Fig 11.

Performance of various lithology categories evaluated by (a) accuracy, (b) precision, (c) recall, and (d) F1−score, respectively. The subplots of each figure are the numerical distribution of each evaluation metric on each label, and the number next to the boxes represent the mean values.

More »

Fig 11 Expand

Fig 12.

Confusion matrix of ResNeSt-50 on testing group.

More »

Fig 12 Expand

Fig 13.

Images of mudstone misclassified as siltstone.

More »

Fig 13 Expand

Fig 14.

Comparison of Params, FLOPs of CNN models.

More »

Fig 14 Expand

Table 2.

Some other evaluation metrics were compared in the training process, testing process, and the model itself.

More »

Table 2 Expand

Fig 15.

Grad-CAM visualization of DenseNet-161 making correct predictions on testing images.

More »

Fig 15 Expand