[1]武明虎,雷常鼎,刘 聪. 改进的轻量级YOLOv4输电线路鸟巢检测方法[J].湖北工业大学学报,2023,(2):11-16.
　WU Minghu,LEI Changding,LIU Cong. Improved YOLOv4 Transmission Lines Bird’s Nest Detection Method[J].,2023,(2):11-16.
点击复制

 改进的轻量级YOLOv4输电线路鸟巢检测方法()
分享到：

《湖北工业大学学报》[ISSN:1003-4684/CN:42-1752/Z]

卷:
期数:
2023年第2期
页码:
11-16
栏目:
湖北工业大学学报
出版日期:
2023-04-30

文章信息/Info

Title:
 Improved YOLOv4 Transmission Lines Bird’s Nest Detection Method
文章编号:
1003-4684(2023)02-0011-06
作者:
 武明虎;  雷常鼎;  刘 聪
 湖北工业大学电气与电子工程学院， 湖北 武汉 430068
Author(s):
WU Minghu; LEI Changding;  LIU Cong
 School of Electrical and Electronic Engineering, Hubei Univ. of Tech., Wuhan 430068, China
关键词:
深度学习;  缺陷检测;  YOLOv4;  注意力机制
Keywords:
 ; deep learning;  defect detecting;  YOLOv4;  Attentional mechanism
分类号:
TP391.4
文献标志码:
A
摘要:
 基于深度学习的输电线路缺陷检测方法往往需要昂贵的硬件支持来实现实时和高精度的技术指标要求，难以大规模推广和普及。针对这一问题，提出一种改进的轻量级YOLOv4神经网络模型方法。具体的，为提高检测精度，在特征提取网络中增加了轻量级注意力机制；为提高检测速度，选用轻量级Mobilenet-v3网络作为YOLOv4的骨干网，同时采用了深度可分离卷积代替了传统YOLOv4模型中的多尺度特征金字塔结构普通卷积。实验结果表明，改进的轻量级YOLOv4模型能有效检测出输电线路中的鸟巢，平均检测精度(AP)值达到97.56%，相较于YOLOv4模型提高了1.80%，检测速度达到YOLOv3的45.8倍，是YOLOv4的56.6倍，具有较强的实际应用价值。
Abstract:
 Transmission line defect detection methods based on deep learning often need expensive hardware support to achieve real time and high precision technical requirements, so it is difficult to be popularized on a large scale. To solve this problem, an improved lightweight YOLOv4 neural network model is proposed. Specifically, in order to improve the detection accuracy, a lightweight attention mechanism is added to the feature extraction network. In order to improve the detection speed, lightweight Mobilenet v3 network is selected as the backbone network of YOLOv4, and deep deprivable convolution is adopted to replace the traditional multi scale characteristic pyramid convolution in YOLOv4 model. The experimental results show that the improved lightweight YOLOv4 model can effectively detect nests in transmission lines, and the average detection accuracy (AP) value reaches 97.56%, which is 1.80% higher than that of YOLOv4 model. The detection speed is 45.8 times that of YOLOv3 and 56.6 times that of YOLOv4, which has strong practical application value.

参考文献/References:

[1] 焦红.直升机巡检输电线路图像中防振锤的识别定位[D].大连:大连海事大学, 2011.
[2] 段旺旺,唐鹏,金炜东,等.基于关键区域HOG特征的铁路接触网鸟巢检测[J].中国铁路, 2015(08): 73-77.
[3] LU J, XU X, XIN L, et al. Detection of bird’s nest in high power lines in the vicinity of remote campus based on combination features and cascade classifier[J]. IEEE Access, 2018 (99):1-1.
[4] CHEN M Y, XU C. Bird’s nest detection method on electricity transmission line tower based on deeply convolutional neural networks[C].∥2020 IEEE 4th Information Technology, Networking, Electronic and Automation Control Conference (ITNEC). IEEE, 2020.
[5] JIANG H,HUANG W,CHEN J, et al. Detection of bird nests on power line patrol using single shot detector[C].∥2019 Chinese Automation Congress (CAC). IEEE, 2020.
[6] BERG AC, FU CY, SZEGEDY C, et al. SSD：单发多盒检测器：，10.1007/978-3-319-46448-0_2[P]. 2015.
[7] 王纪武,罗海保,鱼鹏飞,等.基于Faster R-CNN的多尺度高压塔鸟巢检测[J].北京交通大学学报,2019,43(05):37-43.
[8] 丁建,黄陆明,朱迪锋,曹浩楠.面向高空塔架鸟巢检测的双尺度YOLOv3网络学习[J].西安理工大学学报,2021,37(02):253-260.
[9] REDMON J, FARHADI A. Yolov3: An incremental improvement[J]. arXiv preprint arXiv，2018，1804：02767.
[10] W DONG, L WU, Q WANG, et al. An automatic detection method of bird’s nest on electric tower based on attention full convolutional neural networks[C].∥2021 4th International Conference on Artificial Intelligence and Big Data (ICAIBD), 2021：304-308, doi: 10.1109/ICAIBD51990.2021,9459088.
[11] HOWARD A.G, ZHU M, CHEN B, et al. MobileNets: Efficient convolutional neural networks for mobile vision applications[EB/OL].(2022-02-10) [2020-07-03]. https:∥arxiv.org/abs/1704.04861.
[12] SANDLER M, HOWARD A, ZHU M, et al. Mobilenetv2: Inverted residuals and linear bottlenecks[C].∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Salt Lake City, USA: IEEE Press,2018:4510-4520. 
[13] HOWARD A, SANDLER M, CHU G, et al. Searching for mobilenetv3[C].∥In Proceedings of the IEEE International Conference on Computer Vision. Seoul, Korea: IEEE Press, 2019:1314-1324.
[14] HOU Q, ZHOU D, FENG J. Coordinate attention for efficient mobile network design[C].∥Proceedings of the IEEE/CVF conference on computer vision and pattern recognition，2021: 13 713-13 722.
[15] JIE H, LI S, GANG S, et al. Squeeze and excitation networks[J]. IEEE transactions on pattern analysis and machine intelligence, 2017, 42(08): 2011-2023.

相似文献/References:

[1]周 浩,周先军,邱书畅.下采样迭代和超分辨率重建的图像风格迁移[J].湖北工业大学学报,2020,(1):25.
　ZHOU Hao,ZHOU Xianjun,QIU Shuchang.Image Style Transfer Algorithm Based on Downsampling Iterative and Super-Resolution Reconstruction[J].,2020,(2):25.
[2]张 宇,程 玉,陈建峡,等. 基于深度学习的在线推荐学习系统设计与开发[J].湖北工业大学学报,2021,(5):64.
　ZHANG Yu,CHENG Yu,CHEN Jianxia,et al. Design and Development of Online Recommendation Learning System Based on Deep Learning[J].,2021,(2):64.
[3]张鹏飞,王淑青,王年涛,等. 基于改进MobileNetV3的PCB裸板缺陷检测[J].湖北工业大学学报,2023,(1):27.
　ZHANG Pengfei,WANG Shuqing,WANG Niantao,et al. PCB Bare Board Defect Detection Based on Improved MobileNetV3[J].,2023,(2):27.
[4]李翱翔,马志艳. 基于热红外图像处理的保温墙体缺陷检测方法[J].湖北工业大学学报,2023,(4):24.
　LI Aoxiang,MA ZhiYan. Defect Detection Method of Thermal Insulation Wall Based on Thermal Infrared Image Processing[J].,2023,(2):24.
[5]董　庚１,王焱清１,２,等.基于 ResNet１８网络的油茶果壳籽分选研究[J].湖北工业大学学报,2023,(5):29.
　DONG Gen,WANG Yanqing,SUN Jiwei,et al.Sorting Model of Camellia Fruit Shells and TeaSeeds Based on ResNet18 Network[J].,2023,(2):29.

备注/Memo

备注/Memo:
［收稿日期］ 2022-02-11
［第一作者］ 武明虎（1975-），男，湖北巴东人，湖北工业大学教授，研究方向为信号与信息处理
更新日期/Last Update: 2023-04-26