Citation: Qiu, Z.; Zhao, Z.; Chen, S.;
Zeng, J.; Huang, Y.; Xiang, B.
Application of an Improved YOLOv5
Algorithm in Real-Time Detection of
Foreign Objects by Ground
Penetrating Radar. Remote Sens. 2022,
14, 1895. https://doi.org/10.3390/
rs14081895
Academic Editor: Susana
Lagüela López
Received: 17 February 2022
Accepted: 12 April 2022
Published: 14 April 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Article
Application of an Improved YOLOv5 Algorithm in Real-Time
Detection of Foreign Objects by Ground Penetrating Radar
Zhi Qiu
1,2
, Zuoxi Zhao
1,2,
*, Shaoji Chen
1,2
, Junyuan Zeng
1,2
, Yuan Huang
1,2
and Borui Xiang
1,2
1
College of Engineering, South China Agricultural University, Guangzhou 510642, China;
qiuzhi@stu.scau.edu.cn (Z.Q.); chen.aoi@stu.scau.edu.cn (S.C.); zengjunyuan@stu.scau.edu.cn (J.Z.);
huang_yuan@stu.scau.edu.cn (Y.H.); xiangborui@stu.scau.edu.cn (B.X.)
2
Ministry of Education Key Technologies and Equipment Laboratory of Agricultural Machinery and
Equipment in South China, South China Agricultural University, Guangzhou 510642, China
* Correspondence: zhao_zuoxi@scau.edu.cn; Tel.: +86-136-0004-9101
Abstract:
Ground penetrating radar (GPR) detection is a popular technology in civil engineering.
Because of its advantages of non-destructive testing (NDT) and high work efficiency, GPR is widely
used to detect hard foreign objects in soil. However, the interpretation of GPR images relies heavily
on the work experience of researchers, which may lead to problems of low detection efficiency and a
high false recognition rate. Therefore, this paper proposes a real-time detection technology of GPR
based on deep learning for the application of soil foreign object detection. In this study, the GPR
image signal is obtained in real time by the GPR instrument and software, and the image signals are
preprocessed to improve the signal-to-noise ratio of the GPR image signals and improve the image
quality. Then, in view of the problem that YOLOv5 poorly detects small targets, this study improves
the problems of false detection and missed detection in real-time GPR detection by improving the
network structure of YOLOv5, adding an attention mechanism, data enhancement, and other means.
Finally, by establishing a regression equation for the position information of the ground penetrating
radar, the precise localization of the foreign matter in the underground soil is realized.
Keywords:
ground penetrating radar (GPR); signal processing; YOLOv5; real-time detection; foreign
object location
1. Introduction
Against the background of the current development of unmanned farms, the automa-
tion of agricultural machinery is the future trend of agricultural development [
1
]. Before
automatic agricultural production, it is necessary to check underground for hard foreign
matter that would be harmful to agricultural machinery and tools, as well as humans and
livestock, especially when using the land for the first time; this is a necessary prerequisite
for the safe operation of agricultural machinery and tools. Obviously, these hard foreign
objects cannot usually be observed by the human eye or cameras [
2
]. To better understand
the characteristics of these foreign objects, researchers have developed an efficient method
to detect hard objects buried in soil in real time.
Because electromagnetic wave detection has the advantages of non-destructive de-
tection, a fast detection process, and high detection accuracy, it is widely used for the
detection of underground foreign objects [
3
,
4
], urban pavement [
5
–
7
], bridge safety [
8
,
9
],
and tunnel cavities [
10
,
11
], as well as in archaeological exploration [
12
,
13
] and other sur-
vey experiments. Researchers can use the characteristic components of common GPR
two-dimensional cross-sectional images, which specifically appear as images similar to
hyperbolic features, allowing researchers to quickly identify whether there is foreign matter
in the underground medium [
14
]. However, raw GPR radar images rarely provide geo-
metric information about buried objects. Moreover, original GPR radar images are often
disturbed by noise, such as experimental noise and reflected waves from other materials on
Remote Sens. 2022, 14, 1895. https://doi.org/10.3390/rs14081895 https://www.mdpi.com/journal/remotesensing
