一种基于模板匹配的FPC缺陷检测算法  

A FPC Defect Detection Algorithm Based on Template Matching

作　　者：陈宏辉 姜研 钟俊健 申飘 CHEN Hong-hui;JIANG Yan;ZHONG Jun-jian;SHEN Piao(Guangdong Songshan Polytechnie College,Shaoguan Guangdong 512126,China)

机构地区：[1]广东松山职业技术学院,广东韶关512126

出　　处：《机电产品开发与创新》2024年第6期189-192,共4页Development & Innovation of Machinery & Electrical Products

基　　金：广东松山职业技术学院校级科研项目(2022KJYB09)。

摘　　要：针对柔性电路板(Flexible Printed Circuit,FPC)在生产线上人工检测存在效率低,出错率高的原因,提出一种基于模板匹配的FPC缺陷检测算法。该算法首先将待测FPC图片与模板图片从RGB空间中分离出各自的R通道图片进行模板匹配,对待测FPC图像进行图像配准以获取金手指所在区域,利用该区域进行二值化、开操作、轮廓提取并筛选,提取金手指轮廓图的上下端的点集在轮廓图上进行直线拟合,形成封闭轮廓,通过对该封闭轮廓进行轮廓面积,轮廓长度,以及轮廓长宽比进行判断,实现对缺陷所在区域的定位。实验结果表明该算法检测效果良好,定位精度高,能够实现对FPC表面缺陷的准确定位。Aiming at the low efficiency and high error rate of manual inspection of Flexible Printed Circuit(FPC)on the production line,an FPC defect detection algorithm based on template matching is proposed.The algorithm first separates the FPC picture to be tested and the template picture from the RGB space of their respective R channel pictures for template matching,performs image registration on the test image to obtain the area where the gold finger is located,and uses this area to perform binarization,open operations,Contour extraction and screening,extract the top and bottom points of the gold finger contour map to fit a straight line on the contour map to form a closed contour,and determine the contour area,contour length,and contour aspect ratio of the closed contour to achieve The location of the defect area.Experimental results show that the algorithm has a good detection effect,high positioning accuracy,and can achieve accurate positioning of FPC surface defects.

关 键 词：FPC 模板匹配 轮廓提取 直线拟合 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]