Japan Society of Civil Engineers 2023 Annual Meeting
Sep 14 - Sep 15, 2023Hiroshima University, Hiroshima Institute of Technology
Japan Society of Civil Engineers 2023 Annual Meeting
Sep 14 - Sep 15, 2023Hiroshima University, Hiroshima Institute of Technology
[CS2-04]An Attempt to Introduce Artificial Intelligence Technology for Classification of Drilling Patterns
*Takeru Akutsu1, Takafumi Kitaoka1, Takuma Hashimoto1, Shuntarou Miyanaga2, Kazuo Sakai2(1. Kansai University, 2. TAISEI Corporation)
Keywords:
mountain tunnel,convolutional neural network,tunnel face,drilling pattern
In a mountain tunnel construction, the automation of explosives handling is required from the safety point of view. Firstly, in the context of the need to estimate drilling patterns and the amount of explosive, this study attempted to classify drilling patterns. Specifically, we examined the possibility of classifying drilling patterns using CNN. As a result of the verification, it is possible to use CNN by increasing the amount of face image through data augmentation, even if the number of usable face images is small or the classification pattern is biased.