长江科学院院报 ›› 2023, Vol. 40 ›› Issue (5): 145-152.DOI: 10.11988/ckyyb.20211276

• 工程安全与灾害防治 • 上一篇    下一篇

一种考虑遮挡物的地铁隧道表面裂缝识别方法

黄远远, 彭潜, 徐兵   

  1. 上海勘测设计研究院有限公司,上海 200434
  • 收稿日期:2021-11-30 修回日期:2022-02-17 出版日期:2023-05-01 发布日期:2023-05-22
  • 作者简介:黄远远(1992-),男,安徽宿州人,工程师,硕士,现主要从事工程监测和科研工作。E-mail:huangyuanyuan@sidri.com
  • 基金资助:
    云南省重大科技专项计划项目(202002AF080003);国家重点研发计划项目(2016YFC0401803)

A Method of Identifying Surface Crack of Subway Tunnel in Consideration of Occlusions

HUANG Yuan-yuan, PENG Qian, XU Bing   

  1. Shanghai Investigation, Design & Research Institution Co., Ltd., Shanghai 200434, China
  • Received:2021-11-30 Revised:2022-02-17 Online:2023-05-01 Published:2023-05-22

摘要: 为了提高地铁隧道表面裂缝识别的精度,基于深度学习和机器视觉技术,提出了一种考虑存在遮挡物情况下的地铁盾构隧道裂缝识别与特征提取算法,利用卷积神经网络实现了基于图像修复技术的裂缝特征提取,从而得到了更准确的裂缝信息。首先提出了一种基于U-net网络的地铁隧道管片表面的遮挡物分割方法,用标记出其位置信息;然后以VGG-16为特征提取网络,并搭建Faster R-CNN网络模型,对隧道管片图像中的裂缝进行定位和标记;最后结合二者识别结果可判断出裂缝与遮挡物之间的关系。针对裂缝被遮挡的情况并结合隧道管片图像的特点,提出了一种可修复被遮挡的裂缝信息的网络模型。结果表明采用该裂缝识别与特征提取方法,修复被遮挡的裂缝信息的准确率可达90%,在实际隧道病害检测与综合评估中具有应用价值。

关键词: 盾构隧道, 表面裂缝识别, 特征提取, 深度学习, 机器视觉技术, 图像修复

Abstract: To enhance the accuracy of identifying surface cracks in subway tunnels, we propose an algorithm for recognizing and extracting crack features in subway shield tunnels taking into account the presence of obstructions based on deep learning and machine vision technology. We employed the convolution neural network to extract the features of cracks using image restoration technology, thereby obtaining accurate information on the cracks. First, we proposed a U-net-based method of segmenting the occlusions on the subway tunnel segment surface, marking the position information by a mask map. Then, we built a Faster R-CNN network model using VGG-16 as the feature extraction network to locate and mark the cracks in the tunnel segment image. By combining the results of the two, we can determine the relationship between the cracks and occlusions. Finally, we proposed the deep learning convolution neural network model in which the occluded crack information can be restored. The trained network model can predict the cracks in the occluded area. The experimental results demonstrate that the accuracy rate of the proposed method reaches 90% in repairing the occluded crack information, indicating its application value in actual tunnel disaster detection and comprehensive evaluation.

Key words: shield tunnel, surface crack identification, feature extraction, deep learning, machine vision technology, image restoration

中图分类号: