摘要
针对进口旧机电产品依靠人工外观检验效率较低、失误率大等问题,提出一种基于图像特征分类统计与权重系数(IFCS-WC)的旧机电产品电路板新旧程度判定方法。该方法首先提取图像特征,并将该特征分类;其次,将特征量与预先建立的统计特征数据库的标准特征进行对比,确定其新旧系数;再次,利用特征的新旧系数和权重系数,得出新旧判定系数;最后,将该新旧判定系数与预设值进行比较,得出新旧程度判定结果。该判定方法通过图像特征量的分类统计,快速建立电路板器件的新旧综合判定模型,能够实现电路板新旧程度的自动、便捷、快速、准确检验,可应用于进口旧机电电路板器件的检测。
The used electromechanical products are usually inspected by artificial vision, which has low efficiency and large errors. This paper presents a circuit board condition judgment method for import of used electromechanical products based on image feature classification statistics and weight coefficient （IFCS-WC）. Firstly, the image features is extracted and classified. Secondly, the features are compared with the statistical characteristic database established beforehand to determine the old and new coefficient. The weight coefficient is used to conclude judgment coefficient. Finally, a judgment determination is made by the judgment coefficient compared with preset value. The method can realize circuit board condition automatic, convenient, rapid and accurate detection.
出处
《自动化与信息工程》
2014年第5期23-26,30,共5页
Automation & Information Engineering
基金
广东出入境检验检疫局科技项目(2013GDK16)
关键词
进口旧机电
图像特征
分类统计
权重系数
新旧判定模型
Import Used Electromechanical Products
Image Feature
Classified Statistic
Weight Coefficient
Old and New Evaluation Model
