陶瓷封装外壳多表面外观检测机构的设计与应用

Design and Application of Multi-Surface Appearance Detection Mechanism for Ceramic Packaging Shell

由于对陶瓷封装外壳的特殊设计需求以及对批次间一致性要求越来越高,人工质检不再能完全满足要求。针对外壳的外形特点、缺陷特点以及生产特点,通过对融合人工智能(AI)检测技术的封装外壳缺陷自动检测算法的研究,设计了多金、缺金、凹/凸点等关键外观缺陷的检测方案。通过开发多表面缺陷检测机构,实现陶瓷封装外壳多个表面缺陷检测的自动化,保证了多表面缺陷均可检测性。该检测机构搭配不同镜头和多角度光源,提高了机器视觉采集能力。相比人工检测,该检测机构的检测能力较高,并实现效率提升2倍以上。该多表面外观检测机构的应用提高了产品检验质量和检验一致性,实现了人工检测向全自动检测的跨越。

Due to the special design requirements of ceramic packaging shell and the increasing requirements for batch to batch consistency,manual quality detection can no longer fully meet the requirements.According to the appearance characteristics,defect characteristics and production characteristics of shell,through the study of the automatic detection algorithm of packaging shell defects with artificial intelligence(AI)detection technology,the detection scheme for the key appearance defects including excessive gold,gold deficiency,concave/convex spot was designed.Through the development of multi-surface defect detection mechanism,the automatic detection of multiple surface defects of ceramic packaging shell was realized,ensuring the detectability of multi-surface defects.The detection mechanism was equipped with different lenses and multi-angle light sources to improve the acquisition capability of machine vision.Compared with manual detection,the detection capacity of the detection mechanism is higher,and the efficiency increases by more than 2 times.The application of the multi-surface appearance detection mechanism improves the quality and consistency of product inspection,and realizes the transition from manual detection to automatic detection.