基于图像处理的PCB检测系统研究

发布时间：2018-03-30 07:42

本文选题：自动对焦　切入点：图像拼接　出处：《昆明理工大学》2017年硕士论文

【摘要】：产业的调整及不断升级对产品的整体质量提出了更高的要求,但在整体的印制电路板(Printed Circuit Board,PCB)的制造中,一般水平的PCB过剩,而高质量高附加值的PCB却供不应求。常见的PCB的缺陷检测方法一般都是人工事后检测,该方法会出现漏检、错检的情况,而且不能及时发现并解决问题,因此,一款精度有保障的PCB质量在线检测设备是必须的。本文以PCB制造企业WK为背景,拟研究PCB裸板运用图像处理的在线检测过程。本文借鉴国内外相关行业产品,相关算法以及行业经验,设计了一套基于图像识别的在线检测系统,以图像识别为手段,以搭建的图像采集硬件系统为基础,对PCB裸板进行在线检测。目的是在采集高质量的图片基础上,进行最准确的图像识别并判断PCB裸板是否符合加工要求。本系统主要包括两大部分:硬件部分为该系统的基础支持,主要由图像采集系统,运动控制系统,照明系统三部分组成,其中图像采集系统采用的是CCD(工业照相机)的线阵照相机;软件系统为本系统的核心部分,该部分的主要功能是对硬件系统进行控制,在采集到图像之后,对图像进行处理并判断。本文涉及到的主要算法有,摄像头自动对焦算法、图像拼接算法、图像预处理算法,圆孔图像识别算法。结合系统硬件本身的条件选择最适合本系统的算法构建了一套基于图像识别的PCB裸板在线监测系统平台。通过对系统主要功能的测试,实现了摄像机的自动对焦,对算法进行优化选择之后的图像拼接准确率得到了提高,图像识别的效率准确度得到了保障。采用高精度图像定位,定位功能更加准确,通过对匹配度的数据显示可以直观的判断加工检测件是否合格。与市场上其他的PCB检测设备进行对比,该系统硬件成本较小,采集图像的效果相对稳定,软件界面简单、操作方便、便于一线工人操作使用,更加的符合我国市场的实际需求。
[Abstract]:The adjustment and continuous upgrading of the industry have put forward higher requirements for the overall quality of the products, but in the overall manufacturing of printed Circuit boards, there is a general level of PCB surplus. But the high quality and high added value PCB is in short supply. The common PCB defect detection method is usually manual afterwards detection, this method will have the situation of missing check, wrong check, and can't find and solve the problem in time, so, It is necessary to develop an online PCB quality testing equipment with guaranteed precision. This paper takes WK of PCB manufacturing enterprise as the background, and studies the online detection process of using image processing for PCB bare board. This paper draws lessons from related products of domestic and foreign industries. Related algorithms and industry experience, a set of on-line detection system based on image recognition is designed, which takes image recognition as a means and builds an image acquisition hardware system. The purpose of online detection of PCB bare board is to collect high quality pictures. The system mainly includes two parts: the hardware part is the basic support of the system, mainly by the image acquisition system, motion control system, The lighting system consists of three parts, in which the image acquisition system uses the linear camera of CCD (industrial camera), the software system is the core part of the system, and the main function of this part is to control the hardware system. The main algorithms involved in this paper are camera auto-focus algorithm, image mosaic algorithm, image preprocessing algorithm. The algorithm of circular hole image recognition. According to the condition of system hardware itself, this paper constructs a set of online monitoring system platform of PCB bare board based on image recognition, which is the most suitable algorithm for the system. Through the test of the main functions of the system, The automatic focus of the camera is realized, the accuracy of image mosaic is improved after the optimization of the algorithm, the efficiency and accuracy of image recognition are guaranteed, and the positioning function is more accurate by using high-precision image positioning. Compared with other PCB testing equipment in the market, the hardware cost of the system is small, the effect of image acquisition is relatively stable, and the software interface is simple. Easy to operate, easy for first-line workers to operate, more in line with the actual needs of the market in China.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41

【参考文献】