基于机器视觉的包装盒冲压缺陷检测方法研究
Research on machine vision based detection method for stamping defects in packaging boxes
李明华 1袁嫣红 1彭来湖 1周可 1徐布都2
作者信息
- 1. 浙江理工大学机械工程学院,浙江 杭州 310018
- 2. 浙江理工大学龙港研究院
- 折叠
摘要
针对保健品包装盒在冲压环节产生的边框塑性变形缺陷,构建了一种基于机器视觉的金属包装盒缺陷检测方法.针对盒体棱边区域产生的小范围高光现象,提出一种基于线性拟合的修正算法.针对包装盒的冲压缺陷,采用形状模板匹配算法初步校验并定位,再裁剪出边框内外圈区域,校验该区域灰度值的标准差均值等指标.结果表明:该方法检测准确率在95%以上,平均检测时间小于100ms,能满足实际应用中在线检测的需求.
Abstract
Aiming at the plastic deformation defects occurred during the stamping process of health product packaging boxes,a machine vision-based defect detection method for metal packaging boxes is developed.To address the issue of small-scale highlights generated in the edge regions of the box body,a correction algorithm based on linear fitting is proposed.For the stamping defects of the packaging box,a shape template matching algorithm is used for preliminary verification and localization.The inner and outer ring regions within the frame are then cropped,and indicators such as the standard deviation and mean of the grayscale values in this region are verified.The results show that the accuracy of this method reaches over 95%,with an average detection time of less than 100ms,meeting the requirements for online detection in practical applications.
关键词
机器视觉/保健品包装盒/高光修正/缺陷检测Key words
machine vision/health product packaging boxes/highlight correction/defect detection引用本文复制引用
基金项目
浙江省科技计划项目(2022C01065)
出版年
2023
NETL
NSTL
维普
万方数据