Deep learning based classification of rock structure of tunnel face
Jiayao Chen,Tongjun Yang,Dongming Zhang,Hongwei Huang,Yu Tian
DOI: https://doi.org/10.1016/j.gsf.2020.04.003
IF: 8.9
2021-01-01
Geoscience Frontiers
Abstract:A database including 42400 images of tunnel face over 150 sections are collected from a highway tunnel project in southwest of China.A framework for classifying multiple rock structures from the images of tunnel face is proposed by using the convolutional neural networks (CNN).The framework is able to interpret 5 types of rock structures including mosaic, granular, layered, block, and fragmentation structures with high efficiency and accuracy.The automated interpretation of rock structure can improve the efficiency, accuracy, and consistency of the geological risk assessment of tunnel face. Because of the high uncertainties in the geological images as a result of different regional rock types, as well as in-situ conditions (e.g., temperature, humidity, and construction procedure), previous automated methods have limited performance in classification of rock structure of tunnel face during construction. This paper presents a framework for classifying multiple rock structures based on the geological images of tunnel face using convolutional neural networks (CNN), namely Inception-ResNet-V2 (IRV2). A prototype recognition system is implemented to classify 5 types of rock structures including mosaic, granular, layered, block, and fragmentation structures. The proposed IRV2 network is trained by over 35,000 out of 42,400 images extracted from over 150 sections of tunnel faces and tested by the remaining 7400 images. Furthermore, different hyperparameters of the CNN model are introduced to optimize the most efficient algorithm parameter. Among all the discussed models, i.e., ResNet-50, ResNet-101, and Inception-v4, Inception-ResNet-V2 exhibits the best performance in terms of various indicators, such as precision, recall, F-score, and testing time per image. Meanwhile, the model trained by a large database can obtain the object features more comprehensively, leading to higher accuracy. Compared with the original image classification method, the sub-image method is closer to the reality considering both the accuracy and the perspective of error divergence. The experimental results reveal that the proposed method is optimal and efficient for automated classification of rock structure using the geological images of the tunnel face.
geosciences, multidisciplinary