It is significant to identify rock-mineral microscopic images in geological engineering. The task of microscopic mineral image identification, which is often conducted in the lab, is tedious and time-consuming. Deep learning and convolutional neural networks (CNNs) provide a method to analyze mineral microscopic images efficiently and smartly. In this research, the transfer learning model of mineral microscopic images is established based on Inception-v3 architecture. The four mineral image features, including K-feldspar (Kf), perthite (Pe), plagioclase (Pl), and quartz (Qz or Q), are extracted using Inception-v3. Based on the features, the machine learning methods, logistic regression (LR), support vector machine (SVM), random forest (RF), k-nearest neighbors (KNN), multilayer perceptron (MLP), and gaussian naive Bayes (GNB), are adopted to establish the identification models. The results are evaluated using 10-fold cross-validation. LR, SVM, and MLP have a significant performance among all the models, with accuracy of about 90.0%. The evaluation result shows LR, SVM, and MLP are the outstanding single models in high-dimensional feature analysis. The three models are also selected as the base models in model stacking. The LR model is also set as the meta classifier in the final prediction. The stacking model can achieve 90.9% accuracy, which is higher than all the single models. The result also shows that model stacking effectively improves model performance.

Rock mineral recognition is a costly and time-consuming task when using traditional methods, during which physical and chemical properties are tested at micro- and macro-scale in the laboratory. As a solution, a comprehensive recognition model of 12 kinds of rock minerals can be utilized, based upon the deep learning and transfer learning algorithms. In the process, the texture features of images are extracted and a color model for rock mineral identification can also be established by the K-means algorithm. Finally, a comprehensive identification model is made by combining the deep learning model and color model. The test results of the comprehensive model reveal that color and texture are important features in rock mineral identification, and that deep learning methods can effectively improve identification accuracy. To prove that the comprehensive model could extract effective features of mineral images, we also established a support vector machine (SVM) model and a random forest (RF) model based on Histogram of Oriented Gradient (HOG) features. The comparison indicates that the comprehensive model has the best performance of all.

It is meaningful to study the geological structures exposed on the Earth’s surface, which is paramount to engineering design and construction. In this research, we used 2206 images with 12 labels to identify geological structures based on the Inception-v3 model. Grayscale and color images were adopted in the model. A convolutional neural network (CNN) model was also built in this research. Meanwhile, K nearest neighbors (KNN), artificial neural network (ANN) and extreme gradient boosting (XGBoost) were applied in geological structures classification based on features extracted by the Open Source Computer Vision Library (OpenCV). Finally, the performances of the five methods were compared and the results indicated that KNN, ANN, and XGBoost had a poor performance, with the accuracy of less than 40.0%. CNN was overfitting. The model trained using transfer learning had a significant effect on a small dataset of geological structure images; and the top-1 and top-3 accuracy of the model reached 83.3% and 90.0%, respectively. This shows that texture is the key feature in this research. Transfer learning based on a deep learning model can extract features of small geological structure data effectively, and it is robust in geological structure image classification.

Geochemical discrimination of basaltic magmatism from different tectonic settings remains an essential part of recognizing the magma generation process within the Earth’s mantle. Discriminating among mid-ocean ridge basalt (MORB), ocean island basalt (OIB) and island arc basalt (IAB) is that matters to geologists because they are the three most concerned basalts. Being a supplement to conventional discrimination diagrams, we attempt to utilize the machine learning algorithm (MLA) for basalt tectonic discrimination. A combined MLA termed swarm optimized neural fuzzy inference system (SONFIS) was presented based on neural fuzzy inference system and particle swarm optimization. Two geochemical datasets of basalts from GEOROC and PetDB served as to test the classification performance of SONFIS. Several typical discrimination diagrams and well-established MLAs were also used for performance comparisons with SONFIS. Results indicated that the classification accuracy of SONFIS for MORB, OIB and IAB in both datasets could reach over 90%, superior to other methods. It also turns out that MLAs had certain advantages in making full use of geochemical characteristics and dealing with datasets containing missing data. Therefore, MLAs provide new research tools other than discrimination diagrams for geologists, and the MLA-based technique is worth extending to tectonic discrimination of other volcanic rocks.

Hammering rocks of different strengths can make different sounds. Geological engineers often use this method to approximate the strengths of rocks in geology surveys. This method is quick and convenient but subjective. Inspired by this problem, we present a new, non-destructive method for measuring the surface strengths of rocks based on deep neural network (DNN) and spectrogram analysis. All the hammering sounds are transformed into spectrograms firstly, and a clustering algorithm is presented to filter out the outliers of the spectrograms automatically. One of the most advanced image classification DNN, the Inception-ResNet-v2, is then re-trained with the spectrograms. The results show that the training accurate is up to 94.5%. Following this, three regression algorithms, including Support Vector Machine (SVM), K-Nearest Neighbor (KNN), and Random Forest (RF) are adopted to fit the relationship between the outputs of the DNN and the strength values. The tests show that KNN has the highest fitting accuracy, and SVM has the strongest generalization ability. The strengths (represented by rebound values) of almost all the samples can be predicted within an error of [−5, 5]. Overall, the proposed method has great potential in supporting the implementation of efficient rock strength measurement methods in the field.