基于图像处理的PCB焊点定位研究

发布时间：2018-03-09 10:46

本文选题：PCB　切入点：图像处理　出处：《武汉工程大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着电子科技水平的不断提高和消费电子行业的快速发展,作为电子制造业的支柱行业,PCB板的市场需求也日益增加,该行业制造工艺的复杂化、微型化和密集化的发展趋势,对PCB的生产工艺提出了更高标准的要求。焊点圆盘的定位精度与PCB的贴装质量和成品率这两个关键生产指标密切相关,因此实现上述目标区域的精确定位显得尤为必要。与传统的定位方法相比,基于计算机视觉的AOI系统以其高精度、高效率和无损检测的优势,成为PCB检测定位的主流途径。设计良好的图像处理算法,是基于计算机视觉的PCB焊点圆盘检测定位系统的核心和关键环节,本文对此开展了深入研究,主要内容如下:(1)介绍了PCB行业的发展趋势和焊点圆盘精确快速定位的意义,对当前国内外的相关研究现状进行了总结。介绍基于计算机视觉的PCB焊点圆盘检测定位系统的构成及其各个单元模块的功能。(2)鉴于图像处理模块在定位系统中的核心价值和关键作用,对本文涉及到的图像处理技术进行了一定的研究。针对PCB板的独特构成和特性,设计了一套预处理算法:首先进行图像倾斜度校正,然后进行图像去噪、图像增强和选取R通道显示,用以增强目标区域和背景区域的差异,通过数学形态学运算进行连通区域标定和删除,得到滤除冗余背景信息突出焊点圆盘目标区域的图像。(3)根据目标区域的形状,引入霍夫变换法与最小二乘拟合法两种常见的圆检测方法,通过仿真实验比较其检测定位性能。实验证明,最小二乘拟合法可实现精确的、快速的检测定位,且对由于干扰条件而引起的目标区域部分缺失的情况具有良好的自适应性。
[Abstract]:With the continuous improvement of electronic science and technology and the rapid development of consumer electronics industry, the market demand of PCB board, which is the pillar industry of electronic manufacturing industry, is increasing day by day, and the manufacturing process of this industry is becoming more complicated, miniaturized and intensive. A higher standard is put forward for the production process of PCB. The positioning accuracy of solder joint disc is closely related to the two key production indexes of PCB, that is, the quality of assembly and the rate of finished product. Compared with the traditional localization method, the AOI system based on computer vision has the advantages of high precision, high efficiency and nondestructive testing. The well-designed image processing algorithm is the core and key link of PCB solder spot detection and positioning system based on computer vision. The main contents are as follows: 1) the development trend of PCB industry and the significance of precise and fast positioning of solder joint disc are introduced. This paper summarizes the current research situation at home and abroad. It introduces the structure of PCB solder spot detection and location system based on computer vision and the function of each unit module. 2) in view of the image processing module in the positioning system. Core values and key roles, The image processing technology involved in this paper is studied. According to the unique structure and characteristics of PCB board, a set of preprocessing algorithm is designed: firstly, the image inclination is corrected, then the image denoising is carried out. Image enhancement and selection of R-channel display are used to enhance the difference between the target region and the background region, and the connected region is calibrated and deleted by mathematical morphological operation. The image of eliminating redundant background information to highlight the target area of solder joint disk is obtained. (3) according to the shape of the target area, two common circle detection methods, Hough transform method and least square fitting method, are introduced. The simulation results show that the least square fitting method can achieve accurate and fast detection and localization, and it has good adaptability to the situation of partial absence of target area caused by interference conditions.
【学位授予单位】：武汉工程大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.41;TN41

【相似文献】