Full Text:   <35>

CLC number: 

On-line Access: 2022-07-31

Received: 2022-03-29

Revision Accepted: 2022-07-13

Crosschecked: 0000-00-00

Cited: 0

Clicked: 97

Citations:  Bibtex RefMan EndNote GB/T7714

-   Go to

Article info.
Open peer comments

Journal of Zhejiang University SCIENCE A 1998 Vol.-1 No.-1 P.

http://doi.org/10.1631/jzus.A2200175


Adaptive cropping shallow attention network for defect detection of bridge girder steel using UAV images on railways


Author(s):  Zong-han MU, Yong QIN, Chong-chong YU, Yun-peng WU, Zhi-peng WANG, Huai-zhi YANG, Yong-hui HUANG

Affiliation(s):  Institute State Key Lab of Rail Traffic Control & Safety, Beijing Jiaotong University, Beijing 100091, China; more

Corresponding email(s):   yqin@bjtu.edu.cn

Key Words:  Railway, Bridge, UAV image, Small object detection, Defect detection


Zong-han MU, Yong QIN, Chong-chong YU, Yun-peng WU, Zhi-peng WANG, Huai-zhi YANG, Yong-hui HUANG. Adaptive cropping shallow attention network for defect detection of bridge girder steel using UAV images on railways[J]. Journal of Zhejiang University Science A, 1998, -1(-1): .

@article{title="Adaptive cropping shallow attention network for defect detection of bridge girder steel using UAV images on railways",
author="Zong-han MU, Yong QIN, Chong-chong YU, Yun-peng WU, Zhi-peng WANG, Huai-zhi YANG, Yong-hui HUANG",
journal="Journal of Zhejiang University Science A",
volume="-1",
number="-1",
pages="",
year="1998",
publisher="Zhejiang University Press & Springer",
doi="10.1631/jzus.A2200175"
}

%0 Journal Article
%T Adaptive cropping shallow attention network for defect detection of bridge girder steel using UAV images on railways
%A Zong-han MU
%A Yong QIN
%A Chong-chong YU
%A Yun-peng WU
%A Zhi-peng WANG
%A Huai-zhi YANG
%A Yong-hui HUANG
%J Journal of Zhejiang University SCIENCE A
%V -1
%N -1
%P
%@ 1673-565X
%D 1998
%I Zhejiang University Press & Springer
%DOI 10.1631/jzus.A2200175

TY - JOUR
T1 - Adaptive cropping shallow attention network for defect detection of bridge girder steel using UAV images on railways
A1 - Zong-han MU
A1 - Yong QIN
A1 - Chong-chong YU
A1 - Yun-peng WU
A1 - Zhi-peng WANG
A1 - Huai-zhi YANG
A1 - Yong-hui HUANG
J0 - Journal of Zhejiang University Science A
VL - -1
IS - -1
SP -
EP -
%@ 1673-565X
Y1 - 1998
PB - Zhejiang University Press & Springer
ER -
DOI - 10.1631/jzus.A2200175


Abstract: 
bridges are an important part of railway infrastructure and need regular inspection and maintenance. Using UAV technology to inspect railway infrastructure is an active research issue. However, due to the large size of UAV images, flight distance and height changes, the object scale changes dramatically. At the same time, the elements of interest in railway bridges, such as bolts and corrosion, are small and dense objects, and the sample data set is seriously unbalanced, posing great challenges to the accurate detection of defects. In this paper, an Adaptive Cropping Shallow Attention Network (ACSANet) is proposed, which includes an adaptive cropping strategy for large UAV images and a shallow attention network for small object detection in limited samples. To enhance the accuracy and generalization of the model, the shallow attention network model integrates a CA (Coordinate Attention) attention mechanism module and an α-IOU (Alpha-IOU) loss function, and then carries out defect detection on the bolts, steel surfaces and railings of railway bridges. The test results show that the ACSANet model outperforms the YOLOv5s model using adaptive cropping strategy in terms of the total mAP and missing bolts mAP by 5% and 30%, respectively. Also, compared with the YOLOv5s model that adopts the common cropping strategy, the total mAP and missing bolt mAP are improved by 10% and 60% respectively. Compared with the YOLOv5s model without any cropping strategy, the total mAP and missing bolt mAP are improved by 40% and 67%, respectively.

Darkslateblue:Affiliate; Royal Blue:Author; Turquoise:Article

Open peer comments: Debate/Discuss/Question/Opinion

<1>

Please provide your name, email address and a comment





Journal of Zhejiang University-SCIENCE, 38 Zheda Road, Hangzhou 310027, China
Tel: +86-571-87952783; E-mail: cjzhang@zju.edu.cn
Copyright © 2000 - 2022 Journal of Zhejiang University-SCIENCE