容器外形轮廓智能检测系统的设计  

Design of Intelligent Detection System for Container Contour

作　　者：苟岩岩[1] 姚洪利 GOU Yanyan;YAO Hongli(Department of Mechanical Engineering,Taishan Polytechnic,Taian 271000,China)

机构地区：[1]泰山职业技术学院机电系,山东泰安271000

出　　处：《电子质量》2023年第9期67-71,共5页Electronics Quality

摘　　要：为了解决灌装产品生产过程中长期难以解决的杂瓶干扰问题,设计了一种容器外形轮廓智能检测系统。瓶身外形检测采用准远心光学系统和自学习算法,在保证足够精度指标的前提下,能够将光学结构成本压缩到使用常规远心光学系统的1%,并通过自学习算法自动对容器外形、颜色等信息进行建模。工程应用表明,该系统调试方便、检测稳定、精度高,具有良好的推广应用价值。In order to solve the problem of bottle interference which is difficult to be solved for a long time in the production process of filling products,an intelligent container contour detection system is designed.Quasi telecentric optical system and self-learning algorithm are used for bottle shape detection.On the premise of ensuring sufficient accuracy index,the cost of optical structure can be reduced to 1%of that of conventional telecentric optical system,and the container shape,color and other information can be automatically modeled through self-learning algorithm.The engineering application shows that the system has the advantages of convenient debugging,stable detection,high precision and good popularization and application value.

关 键 词：容器 轮廓检测 机器视觉 自学习算法 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]