基于UNet语义分割模型的矿山采集钻孔裂隙识别研究  

Research on borehole fracture identification based on UNet semantic segmentation model

作　　者：王有志 李缘平 罗甦元 夏一鸣 林伟 WANG You-zhi;LI Yuan-ping;LUO Su-yuan;XIA Yi-ming;LIN Wei(School of Architecture and Civil Engineering,Shenyang University of Technology,Shenyang 110000,China)

机构地区：[1]沈阳工业大学建筑与土木学院,辽宁沈阳110000

出　　处：《世界有色金属》2024年第23期217-219,共3页World Nonferrous Metals

摘　　要：在地质勘探和岩土工程中,钻孔裂隙的准确识别对于评估岩石稳定性至关重要。传统方法依赖人工观察,不仅效率低下,而且结果易受主观影响。随着深度学习技术的兴起,基于UNet架构的语义分割模型为自动化裂隙识别提供了新思路。本研究基于UNet的模型进行训练,通过学习裂隙与非裂隙区域的特征差异,实现裂隙的精确识别。本文使用精确率、召回率和F1分数等指标来评价该模型的裂隙分割性能。实验结果显示,基于UNet的裂隙识别方法能够准确识别出钻孔,大幅提高了识别的准确率和自动化水平。本研究不仅为岩石裂隙识别提供了高效自动化工具,也证明了深度学习在地质领域应用的广阔前景,为岩土工程和地质勘探的发展贡献了新的技术支持。In geological exploration and geotechnical engineering,the accurate identification of borehole fractures is crucial for assessing rock stability.Traditional methods rely on manual observation,which is not only inefficient,but also the results are susceptible to subjective influence.With the rise of deep learning technology,semantic segmentation models based on UNet architecture provide new ideas for automated fracture identification.In this study,the UNet-based model is trained to achieve accurate identification of fissures by learning the feature differences between fissure and non-fissure regions.In this paper,metrics such as precision rate,recall rate and F1 score are used to evaluate the model's performance in fracture segmentation.The experimental results show that the UNet-based fissure identification method can accurately identify boreholes,which significantly improves the accuracy and automation level of identification.This study not only provides an efficient automated tool for rock fissure identification,but also proves the broad prospect of deep learning application in geological field,which contributes new technical support to the development of geotechnical engineering and geological exploration.

关 键 词：孔壁裂隙识别 语义分割模型 UNet网络 

分 类 号：TP31[自动化与计算机技术—计算机软件与理论]