基于机器视觉与SVM分类器的啤酒瓶盖质量检测  

Beer Bottle Cap Quality Inspection Based on Machine Vision and SVM Classifier

作　　者：张航 ZHANG Hang(Qingdao Beer(Rizhao)Co.,Ltd.,Rizhao,Shandong 276800,China)

机构地区：[1]青岛啤酒(日照)有限公司,山东日照276800

出　　处：《自动化应用》2025年第6期9-11,共3页Automation Application

摘　　要：传统的质量检测方式存在效率低下、速度慢、劳动强度大且检测精度不稳定等问题,难以满足现代啤酒生产对高速、高效、高质量的需求,因此,现提出基于机器视觉与SVM分类器的啤酒瓶盖质量检测。首先,基于机器视觉采集图像,并运用图像处理技术来减少或控制图像中的噪声。其次,根据面积、周长、填充率及圆形度等特征提取啤酒瓶盖表面特征信息。最后,基于SVM分类器识别瓶盖缺陷,实现对瓶盖质量的有效检测。实验结果表明,基于机器视觉与SVM分类器的啤酒瓶盖质量检测方法具有较高的质量检测准确率和较快的检测速度,能够有效地识别出瓶盖上的各种缺陷。Traditional quality inspection methods have problems such as low efficiency,slow speed,high labor intensity,and unstable detection accuracy,which are difficult to meet the modern beer production's demand for high speed,efficiency,and quality.Therefore,a beer bottle cap quality inspection based on machine vision and SVM classifier is proposed.Firstly,images are captured based on machine vision and image processing techniques are used to reduce or control noise in the images.Secondly,surface feature information of beer bottle caps is extracted using features such as area,perimeter,filling rate,and circularity.Finally,bottle cap defects are identified using SVM classifier to achieve effective detection of bottle cap quality.The experimental results show that the beer bottle cap quality detection method based on machine vision and SVM classifier has high quality detection accuracy and fast detection speed,and can effectively identify various defects on the bottle cap.

关 键 词：质量检测 啤酒瓶盖 图像处理 SVM分类器 机器视觉 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]