基于YOLO的钢缆表面损坏检测.pdfVIP

计算机系统应用

ISSN

1003-3254,

CODEN

CSAOBNE-mail:

csa@

Computer

Systems

Applications,2024,33(1):134−140

[doi:

10.15888/ki.csa.009364]

©中国科学院软件研究所版权所有.Tel:

+86-10

基于YOLO的钢缆表面损坏检测①

刘际驰,

吕后坤,

李伟

(大连工业大学,

大连

116034)

通信作者:

李伟,

E-mail:

leeway0504@

摘要:

为了解决检测钢缆表面损坏时检测设备资源有限、时间过长等问题,

将深度学习的先进技术以及卷积神经网

络(CNN)应用于钢缆表面损坏检测.

提出了一种基于YOLO的缺陷检测网络模型,

将GhostNet融入主干网络,

并基

于ShuffleNet及注意力机制提出了新的特征提取模块(ShuffleC3),

再对Head部分进行剪枝改进.

实验结果表明,

改

进后网络相比基线YOLOv5s的平均精度提高1.1%,

参数量和计算量分别降低了43.4%和31%,

模型大小减少了

42.3%.

可以在降低网络计算成本的同时,

保持较高的识别精确度,

更好地满足了对钢缆材料表面损坏检测的要求.

关键词:

深度学习;

卷积神经网络;

YOLO;

钢缆;

注意力机制;

表面损坏检测;

目标检测

引用格式:

刘际驰,吕后坤,李伟.基于YOLO的钢缆表面损坏检测.计算机系统应用,2024,33(1):134–140.

/1003-

3254/9364.html

YOLO-basedSurfaceDamageDetectionofSteelCables

LIU

Ji-Chi,

LYU

Hou-Kun,

LI

Wei

(Dalian

Polytechnic

University,

Dalian

116034,

China)

Abstract:

To

solve

the

limited

resources

and

long

time

of

detection

equipment

in

detecting

surface

damage

of

steel

cables,

this

study

applies

advanced

technology

of

deep

learning

and

convolutional

neural

networks

(CNNs)

to

surface

damage

detection

of

the

cables.

On

this

basis,

it

proposes

a

YOLO-based

defect

detection

network

model

to

integrate

GhostNet

into

the

backbone

network,

and

a

new

feature

extraction

module

(ShuffleC3)

based

on

ShuffleNet

and

attention

mechanism,

and

then

prunes

and

improves

the

Head

part.

Experimental

results

show

that

compared

with

the

baseline

YOLOv5s,

the

average

accuracy

of

the

improved

network

is

increased

by

1.1%.

In

addition,

the

number

of

parameters

and

calculations

are

reduced

by

43.4%

and

31%

respectively,

and

the

model

size

is

reduced

by

42.3%.

Thus,

the

proposed

model

can

reduce

the

network

computing

cost

and

maintain

higher

identification

accuracy,

which

better

meets

the