摘要
针对盾构隧道渗漏检测中传统卷积神经网络易造成特征失真、难以高效处理稀疏点云数据的问题,本文提出了一种基于稀疏卷积的隧道渗漏检测方法(Sparse U-Net)。该方法基于三维激光雷达点云数据,通过数据体素化、构建哈希表和规则表,实现稀疏卷积的高效计算,有效捕捉渗漏区域的线状特征。此外,采用编码器-解码器结构精确分割渗漏区域,并引入Focal Loss处理数据不平衡问题。试验表明,该方法在检测精度和资源利用率上均优于传统方法,IoU和准确率分别提高了5.52和3.41个百分点,为盾构隧道渗漏检测提供了高效可靠的技术方案。
Traditional convolutional neural networks face challenges in accurately detecting leakage in shield tunnels,particularly due to feature distortion and inefficiencies in handling sparse point cloud data.To address this,we propose a Sparse U-Net based leakage detection method,leveraging three-dimensional LiDAR point clouds.This method incorporates voxelization,hash table and rule-based sparse convolution operations to efficiently capture linear leakage features.An encoder-decoder architecture is employed for precise leakage segmentation,and Focal Loss is introduced to address class imbalance.Experimental results demonstrate the proposed method significantly improves both accuracy and computational efficiency,achieving increases of 5.52%in IoU and 3.41%in accuracy compared to traditional methods,providing an efficient and reliable solution for leakage detection in shield tunnels.
作者
孙泽信
张安银
段举举
姜俊狄
沈月千
王益波
SUN Zexin;ZHANG Anyin;DUAN Juju;JIANG Jundi;SHEN Yueqian;WANG Yibo(Geo-engineering Investigation Institute of Jiangsu Province,Nanjing 210018,China;School of Earth Sciences and Technology,Hohai University,Nanjing 211100,China;Shaoxing Shangyu District Hangyong Canal Lock Operation Co.,Ltd.,Shaoxing 312300,China)
出处
《测绘通报》
北大核心
2025年第8期137-141,共5页
Bulletin of Surveying and Mapping
基金
国家自然科学基金(41801379)。
关键词
盾构隧道
渗漏检测
稀疏卷积
激光雷达
shield tunnel
leakage detection
sparse convolution
LiDAR
作者简介
孙泽信(1983-),男,硕士,正高级工程师,主要从事精密工程测量研究工作。E-mail:271258150@qq.com;通信作者:沈月千。E-mail:Y.shen_lidar@hhu.edu.cn。
