YOLOv7在探地雷达B-Scan图像解译中的应用

doi:10.13474/j.cnki.11-2246.2023.0227

测绘通报 ›› 2023, Vol. 0 ›› Issue (8): 29-33.doi: 10.13474/j.cnki.11-2246.2023.0227

YOLOv7在探地雷达B-Scan图像解译中的应用

胡荣明, 李鑫, 竞霞, 武建强, 魏青博

西安科技大学测绘科学与技术学院, 陕西西安 710054

收稿日期:2022-11-28 发布日期:2023-09-01
作者简介:胡荣明(1969-),男,博士,教授,主要从事测绘遥感工程领域方面的工作和研究。E-mail:Rmhu2007@163.com
基金资助:
国家自然科学基金(42171394)

Application of YOLOv7 in GPR B-Scan image interpretation

HU Rongming, LI Xin, JING Xia, WU Jianqiang, WEI Qingbo

College of Geomatics Science and Technology, Xi'an University of Science and Technology, Xi'an 710054, China

Received:2022-11-28 Published:2023-09-01

摘要/Abstract

摘要： 探地雷达技术已被广泛应用于隧道衬砌病害检测方面,针对探地雷达B-Scan图像解译,近年来涌现出了许多深度学习算法,但YOLOv7算法在该领域尚未被应用。本文首先通过FDTD正演模拟了隧道衬砌的空洞和渗漏水病害,并实地采集隧道内12条测线数据分析其病害分布情况,以此构成隧道衬砌病害数据集;然后基于YOLOv7算法,利用两类病害的不同图像特征,实现对隧道衬砌病害的自动解译。研究结果表明,利用YOLOv7目标检测算法,对整体病害的识别准确率、召回率分别达到97.87%、90.61%。当IoU取0.5时,对空洞病害和渗漏水病害的识别准确率分别为97.2%和96.4%;最后随机选取100张图像进行测试,测试结果的准确率达94%。最终的试验识别结果能很好地应用在生产项目中。

关键词: YOLOv7, 隧道衬砌病害, 探地雷达, 深度学习, FDTD正演模拟

Abstract: Ground penetrating radar technology has been widely used in tunnel lining disease detection, for ground penetrating radar B-Scan image interpretation, many deep learning algorithms have emerged in recent years, but YOLOv7 algorithm has not been applied in this field. In this paper, the cavities and leakage water diseases of tunnel lining are simulated through FDTD forward performance, and 12 measuring lines in the tunnel are collected on the spot to analyze their disease distribution, so as to form the tunnel lining disease dataset, and then, based on the YOLOv7 algorithm, different image features of the two types of diseases are used to realize the automatic interpretation of tunnel lining diseases. The results show that the target detection algorithm of YOLOv7 is used to identify the overall disease with accuracy and recall rate of 97.87% and 90.61%, respectively. When the threshold IoU is 0.5, the identification accuracy of cavitation disease and leakage water disease is 97.2% and 96.4%, respectively. Finally, 100 images were randomly selected for testing, and the accuracy of the test results reached 94%. The final experimental identification results can be well used in production projects.

Key words: YOLOv7, Tunnel lining diseases, ground penetrating radar(GPR), deep learning, FDTD forward simulations

中图分类号:

胡荣明, 李鑫, 竞霞, 武建强, 魏青博. YOLOv7在探地雷达B-Scan图像解译中的应用[J]. 测绘通报, 2023, 0(8): 29-33.

HU Rongming, LI Xin, JING Xia, WU Jianqiang, WEI Qingbo. Application of YOLOv7 in GPR B-Scan image interpretation[J]. Bulletin of Surveying and Mapping, 2023, 0(8): 29-33.

参考文献

[1] 李尧,李术才,徐磊,等.隧道衬砌病害地质雷达探测正演模拟与应用[J].岩土力学,2016,37(12):3627-3634.
[2] GUO Haonan,SHI Qian,DU Bo,et al.Scene-driven multitask parallel attention network for building extraction in high-resolution remote sensing images[J].IEEE Transactions on Geoscience and Remote Sensing,2021,59(5):4287-4306.
[3] 邵延华,张铎,楚红雨,等.基于深度学习的YOLO目标检测综述[J].电子与信息学报,2022,44(10):3697-3708.
[4] 袁珑,李秀梅,潘振雄,等.面向目标检测的对抗样本综述[J].中国图象图形学报,2022,27(10):2873-2896.
[5] 张东昊,覃晖.基于探地雷达和深度学习的隧道初期支护检测方法[J].现代隧道技术,2020,57(S1):174-178.
[6] QIN Hui,ZHANG Donghao,TANG Yu,et al.Automatic recognition of tunnel lining elements from GPR images using deep convolutional networks with data augmentation[J].Automation in Construction,2021,130:103830.
[7] ZHAO Shuai,SHADABFAR M,ZHANG Dongming,et al.Deep learning-based classification and instance segmentation of leakage-area and scaling images of shield tunnel linings[J].Structural Control and Health Monitoring,2021,28(6):e2732.1-e2732.22.
[8] 任松,朱倩雯,涂歆玥,等.基于深度学习的公路隧道衬砌病害识别方法[J].浙江大学学报(工学版),2022,56(1):92-99.
[9] 彭雨诺,刘敏,万智,等.基于改进YOLO的双网络桥梁表观病害快速检测算法[J].自动化学报,2022,48(4):1018-1032.
[10] 朱家松,郑澳,雷占占,等.基于改进Yolov5的地铁隧道附属设施与衬砌表观病害检测方法[J].铁道科学与工程学报,2023,20(3):1008-1019.
[11] 冯德山,杨子龙.基于深度学习的隧道衬砌结构物探地雷达图像自动识别[J].地球物理学进展,2020,35(4):1552-1556.
[12] 周中,闫龙宾,张俊杰,等.基于YOLOX-G算法的隧道裂缝实时检测[J/OL].铁道科学与工程学报:1-12[2023-07-17].https://doi.org/10.19713/j.cnki.43-1423/u.T20221523.
[13] FEDIUK A,WUNDERLICH T,WILKEN D,et al.Ground penetrating radar measurements in shallow water environments-a case study[J].Remote Sensing,2022,14(15):3659.
[14] WANG C Y,BOCHKOVSKIY A,LIAO H Y M.YOLOv7:Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[EB/OL].[2022-10-18].https://arxiv.org/abs/2207.02696.

[1]	周宝定, 谢沛瑶, 郭文浩, 毛庆洲. 基于激光点云灰度图像的隧道渗水病害检测[J]. 测绘通报, 2023, 0(8): 34-39,90.
[2]	张戬, 高雅. 深度学习遥感影像解译技术在耕地保护中的应用[J]. 测绘通报, 2023, 0(8): 142-145.
[3]	杨盈, 邱芹军, 谢忠, 田苗, 郑诗语, 郑帅. 人在回路学习增强的地理命名实体识别[J]. 测绘通报, 2023, 0(8): 155-160,177.
[4]	钟静, 孙杰, 赖祖龙, 谌一夫. ICESat-2与Sentinel-2数据融合的深度学习浅滩水深测量[J]. 测绘通报, 2023, 0(7): 39-43.
[5]	沈鑫甦, 嵇灵. 面向遥感地类变化检测的U型深度学习神经网络改进方法[J]. 测绘通报, 2023, 0(6): 93-97,103.
[6]	陈桥驿, 闫宇飞, 黄永芳. 融合面向对象和深度置信网络的农村建筑物信息提取[J]. 测绘通报, 2023, 0(6): 134-137.
[7]	王丽梅, 王延正. 基于高分辨率遥感影像的建筑物提取[J]. 测绘通报, 2023, 0(6): 180-183.
[8]	李心媛, 贺智, 楼桉君, 肖曼. 双线性图卷积网络的环南海地区湿地遥感分类[J]. 测绘通报, 2023, 0(5): 44-50.
[9]	蔡军, 温日强, 江伟, 严娇, 卢丽娟. 传统村落航拍图像中民族建筑识别的DeepLabV3+改进算法[J]. 测绘通报, 2023, 0(4): 49-53.
[10]	齐建伟, 王伟峰, 张乐, 王光彦. 基于改进DeepLabV3+算法的遥感影像建筑物变化检测[J]. 测绘通报, 2023, 0(4): 145-149.
[11]	佘佐明, 申勇智, 宋剑虹, 向玉瑨. 利用经典CNN网络方法构建贵阳市道路要素遥感影像自动提取模型[J]. 测绘通报, 2023, 0(4): 177-182.
[12]	王本礼, 王也, 唐先龙, 董胜光. 遥感影像变化图斑智能化提取平台研发与应用[J]. 测绘通报, 2023, 0(2): 150-154.
[13]	吴真真, 唐超, 杨晓飞. 基于深度学习的视频识别及动态监测技术应用[J]. 测绘通报, 2022, 0(9): 23-28.
[14]	田有良, 樊廷立, 唐超. 面向地铁隧道表面渗漏水的快速检测技术[J]. 测绘通报, 2022, 0(9): 29-33.
[15]	钱建国, 张宇琦, 汤圣君, 王伟玺, 李晓明. 最小割与深度学习联合优化的室内粘连点云分割方法[J]. 测绘通报, 2022, 0(9): 45-51.

YOLOv7在探地雷达B-Scan图像解译中的应用

Application of YOLOv7 in GPR B-Scan image interpretation

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价