基于YOLO神经网络模型的自动喷涂线工件分类识别方法
邓人玮,熊瑞平*,卢文翔,徐毅松,史东繁,杨康
(四川大学 机械工程学院,四川 成都 610065)
摘要:针对自动喷涂生产线提出了一种基于YOLO-V3模型的改进密集型工件识别方法。介绍了自动喷涂生产线YOLO-V3 dense改进算法的总体方案流程,通过数据增强的方式获取训练集,并利用密集型网络方法改进算法。对训练后的模型,采用多角度测试集对比及不同神经网络模型对比,评估改进的神经网络模型。测试结果表明,提出的YOLO-V3 dense算法,纵向比较优于YOLO-V2和具有VGG16网络模型的Faster R-CNN,横向比较优于传统的YOLO-V3模型,因此基于YOLO-V3改进的密集型神经网络更加适合用来检测常见喷涂工件。
关键词:深度神经网络;工件识别;实时检测;YOLO
中图分类号:TP389.1 文献标志码:A doi:10.3969/j.issn.1006-0316.2022.03.011
文章编号:1006-0316 (2022) 03-0065-09
Classification Recognition of Automatic Spray Line Work-PieceBased on YOLO Neural Network Model 
DENG Renwei,XIONG Ruiping,LU Wenxiang,XU Yisong,SHI Dongfan,YANG Kang
( School of Mechanical Engineering, Sichuan University, Chengdu 610065, China )
Abstract:An improved method based on YOLO-V3 model is proposed to solve the problem of intelligent identification of work-piece in automatic spraying production line. This paper introduces the General Scheme flow of the improved YOLO-V3dense algorithm for the structure of automatic spraying production line. The training set is obtained by means of data enhancement, and the YOLO-V3 algorithm is improved by means of intensive network method. After the training, the improved neural network model is evaluated by comparing the multi-angle test set and different neural network models. The test results show that the proposed Yolo-v3dense Algorithm is superior to YOLO-V2 and Faster R-CNN with VGG16 network model in longitudinal comparison and to YOLO-V3 model in transverse comparison. Therefore, the dense neural network improved by YOLO-V3 is more suitable for detecting common spraying work-piece.
Key words:deep neural network;work-piece recognition;real-time;YOLO
———————————————
收稿日期:2021-06-15
基金项目:四川省科技厅重点研发项目:智能涂装产线关键技术的开发与集成(2020YFG0119)
作者简介:邓人玮(1996-),男,四川内江人,硕士研究生,主要研究方向为机械电子,E-mail:963081849@qq.com。*通讯作者:熊瑞平(1970-),男,江西瑞金人,博士,副教授,主要研究方向为工业机器人应用及智能控制技术,E-mail:xiongruiping@163.com。
 

 

设为首页  |  加入收藏    |   免责条款
《机械》杂志版权所有     Copyright©2008-2012 Jixiezazhi.com All Rights Reserved 

  电话:028-85925070    传真:028-85925073    E-mail:jixie@vip.163.com

地址:四川省成都锦江工业开发区墨香路48号   邮编:610063

蜀ICP备08103512号

Powered by PageAdmin CMS