摘要
新能源电池的生产、应用和回收必须依据二维码信息进行溯源管理,因此在新能源电池盒上需标注清晰正确的溯源二维码及按国标标注的明码字符信息。为了确保喷涂二维码图形和明码字符信息的质量和正确性,在进行新能源电池盒装配前需进行质量检测。为此,基于深度学习模型的应用,研究和设计溯源二维码机器视觉检测系统。分析新能源溯源二维码的检测需求,开展机器视觉检测系统相机、镜头和光源等硬件的选型及采集系统设计,采用深度学习技术实现溯源二维码的分类标注和模型创建,采用图像处理软件进行溯源二维码和字符检测校对程序的编写,封装程序并完成了二维码检测用户软件设计。对设计的新能源电池溯源二维码机器视觉检测系统进行测试验证,测试数据显示生产效率达15 EA/min、检测精度达0.01 mm、能精准检测出错误二维码和字符,能够快速、精确地进行电池盒二维码质量检测和分拣,满足企业自动化、高效率检测需求,提高生产效率。
The traceability management in the production,utilization and recycling of new energy batteries should base on QR code information,which requires the precise and accurate labeling of traceability QR codes and alphanumeric identifiers on battery boxes in accordance with national standards.To guarantee the quality and correctness of these printed QR code images and alphanumeric characters,it is essential to conduct quality inspections prior to the assembly of new energy battery boxes.Consequently,a study and design of a traceability QR code machine vision inspection system are undertaken,leveraging the application of deep learning models.The analysis of detection requirements for new energy traceability QR codes is conducted,and the selection of hardware components such as cameras,lenses,and light sources for the machine vision inspection system,along with the design of the acquisition system,is performed.Deep learning techniques are employed to facilitate the classification,labeling,and model development of traceability QR codes.Image processing software is utilized to develop the QR code and character detection and verification algorithms,and the programs are encapsulated to finalize the design of the QR code detection user software.The designed new energy battery traceability QR code machine vision inspection system undergoes testing and validation.The results indicated that the system achieves a production efficiency of 15 EA/min,with a detection accuracy of 0.01 mm,and is capable of accurately identifying incorrect QR codes and characters.This system enables rapid and precise quality inspection and sorting of battery box QR codes,fulfilling the enterprise's requirements for automated and efficient detection,thereby enhancing production efficiency.
作者
龙淑嫔
廖书真
陈胜利
周旭华
陈宁
Long Shupin;Liao Shuzhen;Chen Shengli;Zhou Xuhua;Chen Ning(Heyuan Polytechnic,Heyuan,Guangdong 517000,China)
出处
《机电工程技术》
2024年第11期138-142,152,共6页
Mechanical & Electrical Engineering Technology
基金
2022年度广东省普通高校重点领域专项资金项目(2022ZDZX3080)
2022年河源职业技术学院科技计划项目(2022KY-07)。
关键词
新能源电池
溯源二维码
深度学习
机器视觉
new energy battery
traceability QR code
deep learning
machine vision
