ACF粒子自动检测中的图像采集系统研究

发布时间：2018-01-24 05:55

本文关键词： 液晶显示器各向异性的导电薄膜线阵相机自动调焦算法　出处：《电子科技大学》2016年硕士论文　论文类型：学位论文

【摘要】：液晶显示器(LCD)作为近现代崛起并流行的显示器,在各个领域都被广泛使用。21世纪以后,晶玻接装技术(COG)可以让液晶屏变得更轻量化和薄型化。而在COG技术中,各向异性的导电薄膜(ACF)则是最重要的材料之一,ACF压合的导电通道是否良好,直接关系液晶显示屏的质量是否良好。在国外,虽然已经有部分国家或地区研制出了ACF粒子自动检测机,但是成本太高,检测速度也一般。而在国内,大多数厂家还是使用传统的检测方法,通过人工使用金相显微镜观察,不仅效率低,而且需要大量的人工。于是,研制出拥有独立自主知识产权的ACF粒子自动检测机则有着重大的意义,而作为ACF粒子自动检测机中的基础和核心构成之一,图像采集系统的研制更显得迫在眉睫。本文主要分析并设计了ACF粒子自动检测中的图像采集系统的硬件构成、软件系统整体架构以及模块化设计,并对系统各功能模块进行了测试,主要内容如下:1.首先介绍了系统硬件结构的运动流程,然后设计了系统的软件框架。软件框架采用“4+1”视图法来架构,同时按照各系统功能进行分模块实现;2.设计采集图像模块,主要有线阵相机及其配件(镜头、采集卡、光源)等的选取、PC与PLC的串口通信指令的制定、以及线阵相机自动调焦的研究;3.设计图像处理与显示模块,先简单介绍了数字图像的一些相关的编程技术,然后研究了自动调焦算法,最后分析了图像BMP格式的显示方案,其中细节显示算法解决了超大图像显示缓慢卡顿的问题;4.本文重点研究并设计了线阵相机的自动调焦算法。在综合分析了常用的图像清晰度评价函数后,根据调焦图像中边缘细节比较多的特点,决定采用绝对方差函数作为粗调函数,Brenner函数作为精调函数的方法。这种方法既可以保证精度,又能保证整体计算速度达到要求。本文创新性地在ACF粒子自动检测中使用线阵相机的自动调焦算法。在自动调焦时,本文设计的在改变物距时只采集一张图像的方法,很好地优化了自动调焦的速度,据我们查阅国内外重要数据库的结果,在国内还未见相关报道。本系统的检测速度不仅达到了国际平均水平,其成本还大大降低,更适用于大批量生产和普及,拥有极强的市场竞争力。
[Abstract]:Liquid crystal display (LCD) as a modern and popular display, has been widely used in all fields after the 21st century. Glass bonding technology can make the liquid crystal screen lighter and thinner. In COG technology, anisotropic conductive thin film is one of the most important materials. Whether the conductive channel of ACF is good or not is directly related to the quality of LCD. In foreign countries or regions, although some countries or regions have developed ACF particle automatic detection machine, but the cost is too high. Detection speed is also general. In China, most manufacturers still use the traditional detection method, through the manual use of metallographic microscope observation, not only low efficiency, but also need a large number of manual. It is of great significance to develop an automatic ACF particle detector with independent intellectual property rights, and as one of the basic and core components of ACF particle automatic detection machine. The development of image acquisition system is more urgent. This paper mainly analyzes and designs the hardware structure, software architecture and modular design of image acquisition system in ACF particle automatic detection. The main contents are as follows: 1. Firstly, the movement flow of the hardware structure of the system is introduced. Then the software framework of the system is designed. The software framework is constructed by "41" view method, and is implemented by modules according to the functions of each system. 2. Design the image acquisition module, select the serial communication instruction between PC and PLC, and study the automatic focusing of linear array camera. 3. The image processing and display module is designed. Firstly, some related programming techniques of digital image are introduced briefly, then the auto-focusing algorithm is studied. Finally, the display scheme of image BMP format is analyzed. The detail display algorithm solves the problem of slow display of large image. 4. This paper focuses on the research and design of the automatic focusing algorithm of linear camera. After analyzing the commonly used image definition evaluation function, according to the characteristics of more edge details in the focusing image. The method of using absolute variance function as coarse tuning function and Brenner function as fine tuning function can guarantee accuracy. This paper innovatively uses the automatic focusing algorithm of linear camera in ACF particle automatic detection. The method designed in this paper is to collect only one image when changing the object distance, which optimizes the speed of auto-focusing, according to the results of our reference to the important databases at home and abroad. The detection speed of the system has not only reached the international average level, but also its cost has been greatly reduced, which is more suitable for mass production and popularization, and has a strong market competitiveness.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.41

【参考文献】