基于2-D彩色图像模型的PCB焊点检测方法研究  

Research on Inspection Method of Printed Circuit Board's Solder Joint Based on 2-D Model of Color Image

作　　者：吴福培[1,2] 

机构地区：[1]汕头大学工学院机电系,广东汕头515063 [2]汕头轻工装备研究院,广东汕头515063

出　　处：《电测与仪表》2011年第1期18-20,5,共4页Electrical Measurement & Instrumentation

基　　金：国家自然科学基金(60774019);广东省自然科学基金重点项目(2006D930304001);广东省科技计划重大专项(2009A080201004;2010A080402010);汕头大学科研启动项目(09400109)

摘　　要：提出了一种基于2-D彩色图像模型的PCB焊点检测方法。首先,基于3色LED环形结构光源和彩色数字相机的图像采集子系统获取的彩色PCB图像,分析了片式晶片元器件焊点表面信息与其焊点图像色彩分布规律,建立了片式晶片良品焊点类型的焊点检测模型,其次,给出了反映可接受焊点图像中心感兴趣区域(ROI)色彩分布规律的灰度曲线图。在此基础上,以可接受焊点图像的灰度曲线图为基准,与待检测焊点的灰度曲线拟合,根据拟合程度的高低,检测焊点的质量。实验结果表明,本文所提方法的检测准确率为95.8%,检测时间为7s,在满足较高检测检测率的同时,具有较高的检测速度。An inspection method of printed circuit board＇s solder joint based on 2-D model of color image is presented.Based on the solder joint images,which are acquired by a 3 color LED ring structure illuminator and a 3-CCD color digital camera,the relationship between the surface information of solder joint and the color distributing rules of solder joint image is analyzed,and the solder joint inspection model is built for the acceptable solder joint of chip.Then gray-scale curve graph,which reflects the color distributing rules in the region of interest of acceptable solder joint image,is given.On the basis of this,the gray-scale curve of inspected solder joint is fitted with the reference gray-scale curve,which originated from the acceptable solder joint image,and then the solder joint quality is detected by the fitting value scale.Experimental results show that the success rate is 95.8% and the inspection time is 7s using the proposed algorithm,which indicates that the proposed algorithm can achieve both high success rate and inspecting velocity.

关 键 词：自动光学检测 焊点检测 焊点检测模型 灰度曲线图 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]