摘要
针对目前纸卷纸芯植入塞头工作中存在漏装配问题,本研究从深度学习的角度分析相关技术难点,结合完成车间的工艺要求和塞头装配流程,在VC++开发环境下,采用面向对象的软件开发方法设计了一套基于深度学习的塞头有无检测系统。该系统利用工业相机实时采集和处理图像,通过基于深度学习的方式对纸芯塞头进行非接触式有无检测,并与CT-PIS(长泰智能纸品信息处理系统)通信反馈检测结果,实现了纸卷纸芯塞头有无检测的全自动化。实践证明,该系统可以较好地预防塞头漏装配所产生的一系列问题,达到了良好的检测效果。
Aiming at the problem of missing assembly problem in the current work of inserting the plug into the paper core of the paper roll,this paper studied related technical difficulties from the perspective of deep learning,and used object-oriented software development methods to design a plug presence or absence detection system based on deep learning in the VC++development environment combined with the process requirements of the workshop and the plug assembly process.The system used industrial cameras to collect and operate images in real time,and performed non-contact detection of the presence or absence of the plug based on deep learning,then communicated with CT-PIS(Chaint Intelligent Paper Information Processing System)to feed back the detection results.It realized the fully automatic detection of the presence or absence of the plug of the paper roll.Practice had proved that the system has solved a series of problems caused by missing plug assembly,and achieved good detection results.
作者
程宏
张春磊
翁婷
徐昆昆
陈俊
CHENG Hong;ZHANG Chunlei;WENG Ting;XU Kunkun;CHEN Jun(Chaint Corporation,Changsha,Hu’nan Province,410116)
出处
《中国造纸》
CAS
北大核心
2022年第S01期68-73,共6页
China Pulp & Paper
关键词
目标检测
深度学习
工业相机
纸芯塞头
object detection
deep learning
industrial cameras
plug of paper core
作者简介
程宏,先生,硕士,高级工程师,主要从事智能制造、智能工厂规划与设计、智能生产控制系统和智能仓储物流管理系统的研发工作
