广东工业大学学报 ›› 2020, Vol. 37 ›› Issue (03): 42-48.doi: 10.12052/gdutxb.190128

• • 上一篇    下一篇

铁塔航拍图像中鸟巢的YOLOv3识别研究

钟映春1, 孙思语1, 吕帅2, 罗志勇3, 熊勇良3, 何惠清4   

  1. 1. 广东工业大学 自动化学院, 广东 广州 510006;
    2. 广州市树根互联技术有限公司, 广东 广州 510308;
    3. 广州优飞科技有限公司, 广东 广州 510830;
    4. 国家电网江西省萍乡供电公司, 江西 萍乡 330000
  • 收稿日期:2019-10-17 出版日期:2020-05-12 发布日期:2020-05-12
  • 通信作者: 孙思语(1995-),女,硕士研究生,主要研究方向为深度学习和图像处理, E-mail:1052582894@qq.com E-mail:1052582894@qq.com
  • 作者简介:钟映春(1973-),男,副教授,博士,硕士生导师,主要研究方向为模式识别与图像理解
  • 基金资助:
    广东省自然科学基金资助项目（2018A0303130137）；广东省高性能计算重点实验室开放项目（TH1528）；广东省哲学社会科学规划学科共建项目（GD18XJY05）

Recognition of Bird’s Nest on Transmission Tower in Aerial Image of High-volage Power Line by YOLOv3 Algorithm

Zhong Ying-chun1, Sun Si-yu1, Lyu Shuai2, Luo Zhi-yong3, Xiong Yong-liang3, He Hui-qing4   

  1. 1. School of Automation, Guangdong University of Technology, Guangzhou 510006, China;
    2. iRootech Technology Co. Ltd., Guangzhou 510308, China;
    3. Guangzhou Ufly Technology Co. Ltd., Guangzhou 510080, China;
    4. Pingxiang Power Supply Company of State Grid, Pingxiang 330000, China
  • Received:2019-10-17 Online:2020-05-12 Published:2020-05-12

HTML

PDF (PC)

2938

摘要： 电力铁塔上的鸟巢、风筝等异物会严重影响电力架空输电线路的安全性。无人机在巡检过程中会针对电力铁塔进行专门拍照，检测识别铁塔上是否存在鸟巢等异物。针对经典YOLOv3算法在识别铁塔航拍图像中的鸟巢时存在识别精度不高、识别效率偏低、权重参数规模过大等不足，提出了改进方法。首先，设计了改进算法的总体架构，并构建了图像数据集；其次，分别从预测框的宽高损失函数、预测类别不平衡损失函数和神经网络结构等3个方面对经典YOLOv3算法进行改进。实验结果表明，本文的改进措施切实有效，可以在提高识别精度的同时大幅度减小权重参数规模，且识别效率良好。此外，对YOLOv3的改进方法而言，改进其神经网络结构的效果明显好于其他改进措施，为将来在无人机巡检过程中实现实时检测识别目标物奠定了重要基础。

关键词: 高压电力线巡检, 图像检测, 鸟巢识别, YOLOv3算法, 神经网络

Abstract: The safety of high-voltage power line is usually threatened seriously by the foreign matters such as bird’s nests or kites and so on. The transmission tower is an important part of the system of high-voltage power line. So the UAV (Unmanned Aerial Vehicle) often takes photos of tower especially when inspecting the high-voltage power line and these photos usually have to be analyzed by the classical algorithm of YOLOv3 (You Only Look Once Version 3) whether they contain the foreign matters or not. This research is conducted to improve the classical algorithm of YOLOv3 in order to improve its precision and deficiency and unknown scale of weight parameters. Initially, the structure of improvement is designed and image data set is constructed. Second, the classical algorithm of YOLOv3 is improved from three ways: the width and height of loss functions of prediction box, the unbalanced loss function of prediction type and the network structure of classical algorithm are improved respectively. The experiments show that: the improvements proposed are effective, which improves the average recognition precision and reduces the scale of weight parameters greatly and maintains the efficiency. The result of improving the classical algorithm’s neural network structure is obviously better than other ways, which is probably the main direction to improving the algorithm. The investigation provides important basics to detect the objects on real time of UAV.

Key words: high-voltage power transmission line inspection, image detection, bird's nest recognition, YOLOv3 algorithm, neural network

中图分类号: 

  • TP391.41
[1] 张泽徐. 输电线路大跨越铁塔结构设计[J]. 黑龙江科学, 2018, 9(20): 54-55 ZHANG Z X. Design of large-span tower structure for transmission lines [J]. Heilongjiang Science, 2018, 9(20): 54-55
[2] 王少华, 叶自强. 架空输电线路鸟害故障及其防治技术措施[J]. 高压电器, 2011, 47(2): 61-67 WANG S H, YE Z Q. Analysis of bird damage accidents on overhead transmission lines and prevention techniques [J]. High Voltage Apparatus, 2011, 47(2): 61-67
[3] 王鸿海, 张建军. 浅谈鸟害对架空输电线路威胁及对策[J]. 中小企业管理与科技(下旬刊), 2012(7): 308-309 WANG H H, ZHANG J J. Discussion on the threat of birds to overhead transmission lines and countermeasures [J]. Management & Technology of SME, 2012(7): 308-309
[4] 胡智敏, 李凯, 汤国锋, 等. 一种输电线路无人机“巢-巢”巡检新模式[J]. 江西电力, 2018, 42(12): 13-15 HU Z M, LI K, TANG G F, et al. A new mode of transmission line unmanned aerial vehicle nest-nest inspection [J]. Jiangxi Electric Power, 2018, 42(12): 13-15
[5] 王坚俊, 章玮, 楼华辉, 等. 基于无人机图像的输电线检测方法[J]. 计算机系统应用, 2019, 28(2): 233-239 WANG J J, ZHANG W, LOU H H, et al. Power line detection method based on UAV image [J]. Computer Systems & Applications, 2019, 28(2): 233-239
[6] WU X, YUAN P, PENG Q, et al. Detection of bird nests in overhead catenary system images for high-speed rail [J]. Pattern Recognition, 2016, 51: 242-254
[7] 徐晶, 韩军, 童志刚, 等. 一种无人机图像的铁塔上鸟巢检测方法[J]. 计算机工程与应用, 2017, 53(6): 231-235 XU J, HAN J, TONG Z G, et al. Method for detecting bird’s nest on tower based on UAV image [J]. Computer Engineering and Applications, 2017, 53(6): 231-235
[8] LEI X S, SUI Z H. Intelligent fault detection of high voltage line based on the faster R-CNN [J]. Measurement, 2019, 138: 379-385
[9] 陈旭, 张军, 陈文伟, 等. 卷积网络深度学习算法与实例[J]. 广东工业大学学报, 2017, 34(6): 20-26 CHEN X, ZHANG J, CHEN W W, et al. Convolutional neural network algorithm and case [J]. Journal of Guangdong University of Technology, 2017, 34(6): 20-26
[10] 魏伟航, 马乾力, 高治良, 等. 改进卷积神经网络在图像烟雾检测的应用[J]. 智能计算机与应用, 2019, 9(3): 40-44 WEI W H, MA Q L, GAO Z L, et al. Application of improved convolutional neural network in image smoke detection [J]. Intelligent Computer and Applications, 2019, 9(3): 40-44
[11] 刘梦, 王曦廷, 周璐, 等. 基于深度学习与迁移学习的中医舌象提取识别研究[J]. 中医杂志, 2019, 60(10): 835-840 LIU M, WANG X T, ZHOU L, et al. Study on extraction and recognition of traditional Chinese medicine tongue manifestation: based on deep learning and migration learning [J]. Journal of Traditional Chinese Medicine, 2019, 60(10): 835-840
[12] REDMON J, FARHADI A. Yolo9000: better, faster, stronger[C]//IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, Hawaii: IEEE, 2017: 7263-7271.
[13] 陈胜娣, 魏维, 何冰倩, 等. 基于改进的深度卷积神经网络的人体动作识别方法[J]. 计算机应用研究, 2019, 36(3): 945-949 CHEN S D, WEI W, HE B Q, et al. Action recognition based on improved deep convolutional neural network [J]. Application Research of Computers, 2019, 36(3): 945-949
[14] EUNSEOP L, DAIJIN K. Accurate traffic light detection using deep neural network with focal regression loss [J]. Image and Vision Computing, 2019, 87: 24-36
[15] HUANG G, LIU Z, MAATEN L, et al. Densely connected convolutional networks[C]//IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, Hawaii: IEEE, 2017: 4700-4708.
[16] REN S, HE K, GIRSHICK R, et al. Object detection networks on convolutional feature maps [J]. IEEE transactions on pattern analysis and machine intelligence, 2016, 39(7): 1476-1481
[17] 杨孟军, 苏成悦, 陈静, 等. 基于卷积神经网络的视觉闭环检测研究[J]. 广东工业大学学报, 2018, 35(5): 31-37 YANG M J, SU C Y, CHEN J, et al. Loop closure detection for visual SLAM using convolutional neural networks [J]. Journal of Guangdong University of Technology, 2018, 35(5): 31-37
[1] 谢国波, 林立, 林志毅, 贺笛轩, 文刚. 基于YOLOv4-MP的绝缘子爆裂缺陷检测方法[J]. 广东工业大学学报, 2023, 40(02): 15-21.
[2] 张锐, 吕俊. 基于分离结果信噪比估计与自适应调频网络的单通道语音分离技术[J]. 广东工业大学学报, 2023, 40(02): 45-54.
[3] 邱俊豪, 程志键, 林国怀, 任鸿儒, 鲁仁全. 具有执行器故障的非线性系统指定性能控制[J]. 广东工业大学学报, 2023, 40(02): 55-63.
[4] 陈靖宇, 吕毅. 基于脉冲神经网络的冷链制冷机结霜检测方法[J]. 广东工业大学学报, 2023, 40(01): 29-38.
[5] 叶文权, 李斯, 凌捷. 基于多级残差U-Net的稀疏SPECT图像重建[J]. 广东工业大学学报, 2023, 40(01): 61-67.
[6] 彭美春, 阳晨, 李君平, 叶伟斌, 黄文伟. 基于BP神经网络的车辆碳排放测算研究[J]. 广东工业大学学报, 2023, 40(01): 107-112.
[7] 刘洪伟, 林伟振, 温展明, 陈燕君, 易闽琦. 基于MABM的消费者情感倾向识别模型——以电影评论为例[J]. 广东工业大学学报, 2022, 39(06): 1-9.
[8] 章云, 王晓东. 基于受限样本的深度学习综述与思考[J]. 广东工业大学学报, 2022, 39(05): 1-8.
[9] 彭积广, 肖涵臻. 模型预测控制下多移动机器人的跟踪与避障[J]. 广东工业大学学报, 2022, 39(05): 93-101.
[10] 黎耀东, 任志刚, 吴宗泽. 基于深度神经网络的注塑过程预测控制[J]. 广东工业大学学报, 2022, 39(05): 120-126,136.
[11] 曾江毅, 李志生, 欧耀春, 金宇凯. 季节指数改进的PM2.5质量浓度组合预测模型研究[J]. 广东工业大学学报, 2022, 39(03): 89-94.
[12] Gary Yen, 栗波, 谢胜利. 地球流体动力学模型恢复的长短期记忆网络渐进优化方法[J]. 广东工业大学学报, 2021, 38(06): 1-8.
[13] 郭心德, 丁宏强. 离散制造智能工厂场景的AGV路径规划方法[J]. 广东工业大学学报, 2021, 38(06): 70-76.
[14] 黄剑航, 王振友. 基于特征融合的深度学习目标检测算法研究[J]. 广东工业大学学报, 2021, 38(04): 52-58.
[15] 马少鹏, 梁路, 滕少华. 一种轻量级的高光谱遥感图像分类方法[J]. 广东工业大学学报, 2021, 38(03): 29-35.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2017 《广东工业大学学报》编辑部
地址：广州市东风东路729号广东工业大学 邮编：510090
电话：020-37626653,020-37626139,020-37626396
邮箱：xbzrb@gdut.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发