一种关于PCB铜板表面缺陷检测的AOI设计

发布时间：2018-05-20 15:47

本文选题：图像处理 + 自动光学检测　；参考：《北京工业大学》2016年硕士论文

【摘要】：印刷电路板(Printed circuit board,PCB)广泛应用在电子工业的方方面面,随着PCB产品向着高精度、复杂化、超薄型发展,其质量检测问题一直备受关注,传统的PCB质量检测技术已经无法完全满足发展的需求。如何快捷高效的实现PCB表面的缺陷检测,成为人们讨论的热点。自动光学检测(Automatic Optical Inspection,AOI)作为一种新兴的视觉检测方法应运而生并被越来越广泛地应用到PCB表面缺陷检测上面。AOI技术是结合电子学、数字图像处理、光电检测和计算机技术的机器视觉技术,在精密测试技术领域发展迅速。自动光学检测对于无缺陷生产来说是必不可少的一个过程。因此,对于AOI技术在PCB裸板(即铜板)表面缺陷检测进行研究,并不断提出改进意见具有很深远的研究意义。针对AOI技术在PCB裸板表面缺陷检测的研究,做了以下工作。首先,针对PCB裸板表面缺陷设计一个完整的AOI检测方案,包括图像采集、图像预处理、图像分割、缺陷检测与识别等。其次,针对图像采集模块的关键技术,相机标定问题,阐述了相机标定的理论基础以及现有的标定算法的中心思想和操作流程;从操作的繁简度和标定的精确度给出了各种算法的适用场合以及优劣之处。再次,通过对现有图像分割算法的分析与研究,结合PCBy|板图像所特有的高对比度高分辨率的特点,提出了一种基于灰度直方图统计的双门限图像分割技术。最后,针对PCB裸板表面缺陷在线高速检测的要求,结合线阵相机扫描获取到的PCB裸板图像所拥有的对比度高、边缘清晰的特点,提出一种基于轮廓对比的方法对PCB裸板表面缺陷进行检测。利用这种算法实现对各种常见缺陷包括短路、断路、凸起、少铜等PCB裸板表面缺陷进行检测与识别。实验验证该方案可行,可以高效、准确地实现PCB裸板表面缺陷的在线检测。
[Abstract]:Printed circuit (PCB) is widely used in all aspects of the electronic industry. With the development of high precision, complexity and ultra-thin PCB products, the quality testing problem has been paid more attention to. Traditional PCB quality detection technology can not meet the needs of development. How to quickly and efficiently implement PCB surface defect detection has become a hot topic. Automatic Optical Inspection (AOI) has emerged as a new visual detection method and has been applied to PCB surface defect detection more and more widely. AOI is a combination of electronics and digital image processing. The machine vision technology of photoelectric detection and computer technology has developed rapidly in the field of precision testing technology. Automatic optical detection is an essential process for non-defect production. Therefore, it is of great significance to study the AOI technology in the detection of surface defects of PCB bare sheet (i.e. copper sheet), and put forward suggestions for improvement. Aiming at the research of AOI technology in PCB bare plate surface defect detection, the following work has been done. Firstly, a complete AOI detection scheme is designed for PCB bare plate surface defects, including image acquisition, image preprocessing, image segmentation, defect detection and recognition, and so on. Secondly, aiming at the key technology of image acquisition module, camera calibration, the theoretical basis of camera calibration and the central idea and operation flow of the existing calibration algorithm are expounded. The applications and advantages of various algorithms are given from the complexity of operation and the accuracy of calibration. Thirdly, through the analysis and research of the existing image segmentation algorithm, combining the characteristic of high contrast and high resolution of PCBy board image, a double-threshold image segmentation technology based on gray-scale histogram statistics is proposed. Finally, according to the requirement of on-line high-speed detection of PCB bare plate surface defects, combined with the characteristics of high contrast and clear edge of PCB bare plate images obtained by linear array camera scanning, A method based on contour contrast is proposed to detect the surface defects of PCB bare plate. This algorithm is used to detect and identify the surface defects of PCB bare plate, such as short circuit, open circuit, protruding, less copper and so on. The experimental results show that the scheme is feasible, and can be used to detect the surface defects of PCB bare plate efficiently and accurately.
【学位授予单位】：北京工业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.41;TN41

【相似文献】