基于智能数据增强和改进YOLOv3算法的接触网吊弦及支架检测研究

doi:10.11896/jsjkx.200200053

摘要/Abstract

摘要： 接触网是铁路上空架设的为电力机车供电的输电线路,其支架是铁路牵引供电的关键支撑部件,而接触网吊弦是输送电能的关键部件,一旦出现故障,影响巨大,严重时可能引发弓网事故,从而给列车运行带来安全隐患。找到高效准确定位两个关键设备的方法对后续异常判断具有重要意义。针对此问题,提出了一种基于智能数据的增强算法,随机选取一种或多种数据增强方法来对接触网图片进行增强;并对YOLOv3目标检测算法进行改进,增强特征提取网络,设计5个不同尺度的卷积特征图来构成特征金字塔。将改进算法和数据增强算法相结合,最终实现对接触网的吊弦和支架的检测,使用此算法在测试集上获得了93.5%的mAP(均值平均精度),速度达到45帧/秒,在保持较高精度的情况下实现了对接触网吊弦和支架的实时定位。

关键词: YOLOv3, 吊弦和支架检测, 接触网, 卷积神经网络, 数据增强

Abstract: Catenary is a transmission line over the railway to supply power for electric locomotives,its support and dropper both are the key components of railway power transmission,it will make a huge impact if there is a failure.Catenary accidents may occur in serious cases,could cause hidden danger of High-speed trains.It is of great significance to find an efficient and accurate posi-tioning method of these two equipments to facilitate the subsequent abnormal judgment.This paper focuses on this problem,pre-sents an intelligent data augment algorithm,it can randomly select one or more data augment methods to enhance the catenary picture.In addition,this paper proposes an improved YOLOv3 algorithm,5 groups of feature pyramids with different scales are designed by enhancing feature extraction network.Finally,combining the improved algorithm with the data augment algorithm,to realize dropper and support detection task.The mAP of the algorithm on the test dataset is 93.5%,the recognition rate is 45 fps.This method realizes the real-time detection of dropper and support under the high precision.

Key words: Catenary, Convolutional neural network, Data augmentation, Dropper and support detection, YOLOv3

中图分类号:

TP391

刘舒康, 唐鹏, 金炜东. 基于智能数据增强和改进YOLOv3算法的接触网吊弦及支架检测研究[J]. 计算机科学, 2020, 47(11A): 178-182. https://doi.org/10.11896/jsjkx.200200053

LIU Shu-kang, TANG Peng, JIN Wei-dong. Study on Catenary Dropper and Support Detection Based on Intelligent Data Augmentation and Improved YOLOv3[J]. Computer Science, 2020, 47(11A): 178-182. https://doi.org/10.11896/jsjkx.200200053

参考文献

[1] WANG X S.Research on the influence of tension on catenary safety state and on-line monitoring system[D].Beijing:Beijing Jiaotong University,2014.
[2] WANG J W.Classification and segmentation of fundus imagesbased on deep learning[D].Guangzhou:South China University of Technology,2019.
[3] WANG L,FAN X Y,CHEN J H,et al.3D object detection based on sparse convolution neural network and feature fusion for autonomous driving in smart cities[J].Sustainable Cities and Society,2020,54(3):102002.
[4] LUAN R S,LIU G F,et al.Application of dynamic face recogni-tion in investigation[J].China Criminal Police College,2019(5):122-128.
[5] YANG P.Dual discriminator generated countermeasure networkand its application in OCS nest detection[D].Chengdu:Southwest Jiaotong University,2018.
[6] HE D Q,JIANG Z,CHEN J Y,et al.Research on bird nest detection method of Railway Catenary Based on deep convolution neural network[J].Electric Drive for Locomotives,2019(4):12-16.
[7] ULLMANS.Object recognition and segmentation by a frag-ment-based hierarchy[J].Trends in Cognitive Sciences,2007,11(2):58-64.
[8] PEREZ L,WANG J.The Effectiveness of Data Augmentation in Image Classification using Deep Learning[J].arXiv:1712.04621,2017.
[9] GOODFELLOW I J,POUGET-ABADIE J,MIRZA M,et al.Generative Adversarial Networks[J].Advances in Neural Information Processing Systems,2014,3:2672-2680.
[10] REDMON J,FARHADI A.Yolov3:An incremental improve-ment[J].arXiv:1804.02767,2018.
[11] REDMON J,FARHADI A.YOLO9000:Better,Faster,Stronger[J].arXiv:1612.08242v,12016.
[12] GIRSHICK R.Fast R-CNN[C]//IEEE Conference on Compu-ter Vision and Pattern Recognition.2015
[13] REN S Q,HE K M,GIRSHICK R,et al.Faster R-CNN:to-wards real-time object detection with region proposalnetworks[J].IEEE Transactionson Pattern Analysis andMachine Intelligence,2017,39(6):1137-1149.
[14] KRIZHEVSKY A,SUTSKEVER I,HINTONG E.ImageNetclassification with deep convolutional neural networks[C]//International Conference on Neural Information Processing Systems.Curran Associates Inc,2012:1097-1105.

相关文章 15

[1]	周乐员, 张剑华, 袁甜甜, 陈胜勇. 多层注意力机制融合的序列到序列中国连续手语识别和翻译 Sequence-to-Sequence Chinese Continuous Sign Language Recognition and Translation with Multi- layer Attention Mechanism Fusion 计算机科学, 2022, 49(9): 155-161. https://doi.org/10.11896/jsjkx.210800026
[2]	李宗民, 张玉鹏, 刘玉杰, 李华. 基于可变形图卷积的点云表征学习 Deformable Graph Convolutional Networks Based Point Cloud Representation Learning 计算机科学, 2022, 49(8): 273-278. https://doi.org/10.11896/jsjkx.210900023
[3]	陈泳全, 姜瑛. 基于卷积神经网络的APP用户行为分析方法 Analysis Method of APP User Behavior Based on Convolutional Neural Network 计算机科学, 2022, 49(8): 78-85. https://doi.org/10.11896/jsjkx.210700121
[4]	朱承璋, 黄嘉儿, 肖亚龙, 王晗, 邹北骥. 基于注意力机制的医学影像深度哈希检索算法 Deep Hash Retrieval Algorithm for Medical Images Based on Attention Mechanism 计算机科学, 2022, 49(8): 113-119. https://doi.org/10.11896/jsjkx.210700153
[5]	檀莹莹, 王俊丽, 张超波. 基于图卷积神经网络的文本分类方法研究综述 Review of Text Classification Methods Based on Graph Convolutional Network 计算机科学, 2022, 49(8): 205-216. https://doi.org/10.11896/jsjkx.210800064
[6]	张颖涛, 张杰, 张睿, 张文强. 全局信息引导的真实图像风格迁移 Photorealistic Style Transfer Guided by Global Information 计算机科学, 2022, 49(7): 100-105. https://doi.org/10.11896/jsjkx.210600036
[7]	戴朝霞, 李锦欣, 张向东, 徐旭, 梅林, 张亮. 基于DNGAN的磁共振图像超分辨率重建算法 Super-resolution Reconstruction of MRI Based on DNGAN 计算机科学, 2022, 49(7): 113-119. https://doi.org/10.11896/jsjkx.210600105
[8]	刘月红, 牛少华, 神显豪. 基于卷积神经网络的虚拟现实视频帧内预测编码 Virtual Reality Video Intraframe Prediction Coding Based on Convolutional Neural Network 计算机科学, 2022, 49(7): 127-131. https://doi.org/10.11896/jsjkx.211100179
[9]	徐鸣珂, 张帆. Head Fusion:一种提高语音情绪识别的准确性和鲁棒性的方法 Head Fusion:A Method to Improve Accuracy and Robustness of Speech Emotion Recognition 计算机科学, 2022, 49(7): 132-141. https://doi.org/10.11896/jsjkx.210100085
[10]	金方焱, 王秀利. 融合RACNN和BiLSTM的金融领域事件隐式因果关系抽取 Implicit Causality Extraction of Financial Events Integrating RACNN and BiLSTM 计算机科学, 2022, 49(7): 179-186. https://doi.org/10.11896/jsjkx.210500190
[11]	孙福权, 崔志清, 邹彭, 张琨. 基于多尺度特征的脑肿瘤分割算法 Brain Tumor Segmentation Algorithm Based on Multi-scale Features 计算机科学, 2022, 49(6A): 12-16. https://doi.org/10.11896/jsjkx.210700217
[12]	吴子斌, 闫巧. 基于动量的映射式梯度下降算法 Projected Gradient Descent Algorithm with Momentum 计算机科学, 2022, 49(6A): 178-183. https://doi.org/10.11896/jsjkx.210500039
[13]	张嘉淏, 刘峰, 齐佳音. 一种基于Bottleneck Transformer的轻量级微表情识别架构 Lightweight Micro-expression Recognition Architecture Based on Bottleneck Transformer 计算机科学, 2022, 49(6A): 370-377. https://doi.org/10.11896/jsjkx.210500023
[14]	王建明, 陈响育, 杨自忠, 史晨阳, 张宇航, 钱正坤. 不同数据增强方法对模型识别精度的影响 Influence of Different Data Augmentation Methods on Model Recognition Accuracy 计算机科学, 2022, 49(6A): 418-423. https://doi.org/10.11896/jsjkx.210700210
[15]	孙洁琪, 李亚峰, 张文博, 刘鹏辉. 基于离散小波变换的双域特征融合深度卷积神经网络 Dual-field Feature Fusion Deep Convolutional Neural Network Based on Discrete Wavelet Transformation 计算机科学, 2022, 49(6A): 434-440. https://doi.org/10.11896/jsjkx.210900199

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed