Automatic Recognition of Tunnel Water Leakage Based on Adaptive Information Extraction Network and Multiscale Feature Enhancement Module

Water leakage in metro tunnels is a critical safety indicator, necessitating regular inspections to avert catastrophic failures. Deep learning-based computer vision is currently utilized to detect water leakage in metro tunnels. However, challenges like large model parameters, low detection accuracy...

Full description

Saved in:
Bibliographic Details
Main Authors: Dandan Wang, Gongyu Hou, Qinhuang Chen, Weiyi Li, Haoxiang Li, Yaohua Shao, Xunan Yu
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10804770/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849219995619295232
author Dandan Wang
Gongyu Hou
Qinhuang Chen
Weiyi Li
Haoxiang Li
Yaohua Shao
Xunan Yu
author_facet Dandan Wang
Gongyu Hou
Qinhuang Chen
Weiyi Li
Haoxiang Li
Yaohua Shao
Xunan Yu
author_sort Dandan Wang
collection DOAJ
description Water leakage in metro tunnels is a critical safety indicator, necessitating regular inspections to avert catastrophic failures. Deep learning-based computer vision is currently utilized to detect water leakage in metro tunnels. However, challenges like large model parameters, low detection accuracy, and poor robustness in complex environments persist. Therefore, this study proposes a novel lightweight semantic segmentation model. An adaptive information extraction network, integrating spatial and channel squeeze-and-excitation mechanisms, is adopted in the encoder to enhance critical feature representation and accelerate inference. Additionally, a multiscale and lightweight feature enhancement module is introduced to capture global contextual information while reducing the number of parameters. The decoder employs a sequential up-sampling strategy, complemented by skip connections, to integrate multiscale features effectively and enhance gradient flow. The experimental results show that the proposed model has better segmentation performance than other compared models, which is very suitable for detecting metro tunnel leakage in complex environments.
format Article
id doaj-art-392c789230894f7788e7a1d2c050dda9
institution Kabale University
issn 2169-3536
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-392c789230894f7788e7a1d2c050dda92024-12-25T00:01:39ZengIEEEIEEE Access2169-35362024-01-011219258619260210.1109/ACCESS.2024.351932110804770Automatic Recognition of Tunnel Water Leakage Based on Adaptive Information Extraction Network and Multiscale Feature Enhancement ModuleDandan Wang0https://orcid.org/0009-0005-7197-3576Gongyu Hou1Qinhuang Chen2https://orcid.org/0000-0002-5306-6033Weiyi Li3https://orcid.org/0009-0001-7636-9766Haoxiang Li4https://orcid.org/0000-0002-8838-8227Yaohua Shao5https://orcid.org/0009-0001-2209-6125Xunan Yu6https://orcid.org/0009-0007-7359-5745School of Mechanics and Civil Engineering, China University of Mining and Technology-Beijing, Beijing, ChinaSchool of Mechanics and Civil Engineering, China University of Mining and Technology-Beijing, Beijing, ChinaSchool of Mechanics and Civil Engineering, China University of Mining and Technology-Beijing, Beijing, ChinaSchool of Mechanics and Civil Engineering, China University of Mining and Technology-Beijing, Beijing, ChinaSchool of Mechanics and Civil Engineering, China University of Mining and Technology-Beijing, Beijing, ChinaSchool of Mechanics and Civil Engineering, China University of Mining and Technology-Beijing, Beijing, ChinaSchool of Mechanics and Civil Engineering, China University of Mining and Technology-Beijing, Beijing, ChinaWater leakage in metro tunnels is a critical safety indicator, necessitating regular inspections to avert catastrophic failures. Deep learning-based computer vision is currently utilized to detect water leakage in metro tunnels. However, challenges like large model parameters, low detection accuracy, and poor robustness in complex environments persist. Therefore, this study proposes a novel lightweight semantic segmentation model. An adaptive information extraction network, integrating spatial and channel squeeze-and-excitation mechanisms, is adopted in the encoder to enhance critical feature representation and accelerate inference. Additionally, a multiscale and lightweight feature enhancement module is introduced to capture global contextual information while reducing the number of parameters. The decoder employs a sequential up-sampling strategy, complemented by skip connections, to integrate multiscale features effectively and enhance gradient flow. The experimental results show that the proposed model has better segmentation performance than other compared models, which is very suitable for detecting metro tunnel leakage in complex environments.https://ieeexplore.ieee.org/document/10804770/Deep learninglightweightsemantic segmentationtunnel water leakage
spellingShingle Dandan Wang
Gongyu Hou
Qinhuang Chen
Weiyi Li
Haoxiang Li
Yaohua Shao
Xunan Yu
Automatic Recognition of Tunnel Water Leakage Based on Adaptive Information Extraction Network and Multiscale Feature Enhancement Module
IEEE Access
Deep learning
lightweight
semantic segmentation
tunnel water leakage
title Automatic Recognition of Tunnel Water Leakage Based on Adaptive Information Extraction Network and Multiscale Feature Enhancement Module
title_full Automatic Recognition of Tunnel Water Leakage Based on Adaptive Information Extraction Network and Multiscale Feature Enhancement Module
title_fullStr Automatic Recognition of Tunnel Water Leakage Based on Adaptive Information Extraction Network and Multiscale Feature Enhancement Module
title_full_unstemmed Automatic Recognition of Tunnel Water Leakage Based on Adaptive Information Extraction Network and Multiscale Feature Enhancement Module
title_short Automatic Recognition of Tunnel Water Leakage Based on Adaptive Information Extraction Network and Multiscale Feature Enhancement Module
title_sort automatic recognition of tunnel water leakage based on adaptive information extraction network and multiscale feature enhancement module
topic Deep learning
lightweight
semantic segmentation
tunnel water leakage
url https://ieeexplore.ieee.org/document/10804770/
work_keys_str_mv AT dandanwang automaticrecognitionoftunnelwaterleakagebasedonadaptiveinformationextractionnetworkandmultiscalefeatureenhancementmodule
AT gongyuhou automaticrecognitionoftunnelwaterleakagebasedonadaptiveinformationextractionnetworkandmultiscalefeatureenhancementmodule
AT qinhuangchen automaticrecognitionoftunnelwaterleakagebasedonadaptiveinformationextractionnetworkandmultiscalefeatureenhancementmodule
AT weiyili automaticrecognitionoftunnelwaterleakagebasedonadaptiveinformationextractionnetworkandmultiscalefeatureenhancementmodule
AT haoxiangli automaticrecognitionoftunnelwaterleakagebasedonadaptiveinformationextractionnetworkandmultiscalefeatureenhancementmodule
AT yaohuashao automaticrecognitionoftunnelwaterleakagebasedonadaptiveinformationextractionnetworkandmultiscalefeatureenhancementmodule
AT xunanyu automaticrecognitionoftunnelwaterleakagebasedonadaptiveinformationextractionnetworkandmultiscalefeatureenhancementmodule