基于LabVIEW编程平台实现玻璃缺陷检测

以机器视觉技术为基础,利用LabVIEW建立生产者-消费者模式,调用Halcon图像处理算子获取玻璃图像并进行图像处理,采用灰度变化对图像进行对比增强突出缺陷和使用中值滤波进一步削弱图像噪声,利用统计模式识别算法在LabVIEW编程平台上构建一个基于机器视觉的玻璃缺陷检测系统,该系统能够实现对玻璃缺陷的分析和提取,具有高效的实时性.     

基于图像处理的手机屏幕缺陷检测系统研究

针对手机屏幕生产中的缺陷检测问题,提出了基于机器视觉的手机屏幕玻璃缺陷检测系统,采用光学成像模块、图像采集模块和图像处理软件设计模块对手机屏幕玻璃缺陷进行识别检测,检测效率高,实时性好.对手机屏幕玻璃缺陷检测所需用到的算法进行了研究,针对手机屏幕玻璃缺陷细微、痕浅等特点,提出了基于图像快速匹配的差分算法,实现了缺陷特征的准确提取.     

基于机器视觉的产品外观检测系统设计

分析了机器视觉技术发展现状及未来趋势,设计了一种基于机器视觉技术的产品外观检测系统,通过图像采集模块采集待检产品视觉图像,通过专用控制器对图像进行处理,最终识别出待检产品外观缺陷。     

基于机器视觉的液晶基板表面缺陷检测系统

介绍一种基于机器视觉的液晶玻璃基板质量在线检测系统。利用分布式视觉处理技术、采用模块化的图像处理系统设计,能够实现缺陷的精确提取与对缺陷的智能分类和分级,满足LCD液晶玻璃基板质量在线检测的需要。     

低温型风力发电机叶片缺陷自动检测方法研究

针对人工检测发电机叶片缺陷时存在的检测结果不精准,检测时间较长的问题,以低温型风力发电机为研究对象,设计了一种基于机器视觉的叶片缺陷自动检测方法。在分析叶片缺陷特征及参数的前提下生成叶片缺陷图像,利用数字图像处理技术对叶片缺陷图像进行灰度化与滤波处理,采用通域递归法提取叶片的缺陷特征并构建最优阈值,最后利用机器视觉技术,实现发电机叶片缺陷的自动检测。在MATLAB平台内模拟发电机叶片缺陷的检测过程,结果显示,基于机器视觉的检测方法能够准确检测出发电机叶片表面的擦痕、裂纹等缺陷,较人工检测精准度高26.7%,且检测耗时较短,说明机器检测具备有效性。     

机器视觉技术在药液灌装量检测中的应用

药品生产质量的检测是非常重要的。针对玻璃安瓿瓶药液灌装量的检测,提出了一种基于机器视觉技术的、适于实时在线检测和控制的方法,即通过计算机视觉系统,采集水针剂玻璃安瓿瓶图像,利用机器视觉软件HALCON的灰度值形态学算法分割图像,然后查找边界跳变点、进行液位测量,最终作出判断决策。实验研究证明该方法准确度高、检测结果客观等,具有十分明显的优势。     

机器视觉表面缺陷检测综述

目的 工业产品的表面缺陷对产品的美观度、舒适度和使用性能等带来不良影响,所以生产企业对产品的表面缺陷进行检测以便及时发现并加以控制。机器视觉的检测方法可以很大程度上克服人工检测方法的抽检率低、准确性不高、实时性差、效率低、劳动强度大等弊端,在现代工业中得到越来越广泛的研究和应用。方法 以机器视觉表面缺陷检测为研究对象,在广泛调研相关文献和发展成果的基础上,对基于机器视觉在表面缺陷检测领域的应用进行了综述。分析了典型机器视觉表面缺陷检测系统的工作原理和基本结构,阐述了表面缺陷视觉检测的研究现状、现有视觉软件和硬件平台,综述了机器视觉检测所涉及到的图像预处理算法、图像分割算法、图像特征提取及其选择算法、图像识别等相关理论和算法研究,并对每种主要方法的基本思想、特点和存在的局限性进行了总结,对未来可能的发展方向进行展望。结果 机器视觉表面缺陷检测系统中,图像处理和分析算法是重要内容,算法各有优缺点和其适应范围。如何提高算法的准确性、实时性和鲁棒性,一直是研究者们努力的方向。结论 机器视觉是对人类视觉的模拟,机器视觉表面检测涉及众多学科和理论,如何使检测进一步向自动化和智能化方向发展,还需要更深入的研究。     

基于机器视觉的纺织品缺陷检测系统的研究与实现

布匹缺陷检测是纺织行业生产过程中布匹质量得以保障的重要环节,机器视觉技术的发展使得利用数字图像处理技术来检测布匹缺陷成为可能。设计通过分析摄像机采集的布匹纹理特征,应用机器视觉技术判别出布匹上的瑕疵,并进行布匹瑕疵数据分类管理。摄像机实时采集的布匹图像经过ROI提取、图像预处理、图像增强、图像分割、灰度图像二值化、提取特征区域轮廓等操作即可判别布匹中所含缺陷。     

机器视觉在钢球表面缺陷检测中的应用

提出一种基于机器视觉钢球表面缺陷识别方法, 利用计算机图像技术采集钢球表面图像信号, 采用图像比对法对图像信号进行缺陷识别分析.实践表明使用本文方法能够实现钢球表面缺陷的自动检测,具有可靠、高效的特点.     

基于视觉传感器的PCB缺陷检测系统的研究与实现

为了实现PCB缺陷的在线自动检测,设计了一种PCB缺陷自动检测系统,该系统主要由机器臂、电气控制系统以及视觉传感器系统等组成。通过可编程控制的图像采集系统获取高质量的原始视觉图像,利用图像处理实现对缺陷目标的自动检测及识别。实验结果验证了该系统检测PCB板缺陷的高效性和实时性。     

基于激光反射投影原理的重叠玻璃尺寸测量方法研究

论文在激光投影法的基础上,结合重叠玻璃的投影特性提出了一种新的测量重叠玻璃平面尺寸的方法---基于激光反射投影原理的移动式测量方法,解决了使用非接触测量法测量重叠玻璃平面尺寸存在的难题。该方法通过获取经玻璃反射后的激光投影线段图像,分析被反射后激光线的投影特性,提取了反映玻璃平面尺寸的特征,并利用该特征分析计算出了重叠玻璃的平面尺寸。论文根据此方法设计了一套重叠玻璃平面尺寸测量实验系统。实验数据表明：这个测量方法能达到生产的精度要求,对于重叠玻璃的平面尺寸测量具有可行性,可应用于流水线上的重叠玻璃平面尺寸在线测量。     

Effect of mold stiffness on surface flatness of mold-pressed glass

This paper presents the effect of mold stiffness on the surface flatness of mold-pressed glass for glass molding and imprint processes. Phosphate and borosilicate glasses were pressed at their viscoelastic temperature region using flat molds of several kinds and subsequently cooled to room temperature. Surface flatness of the mold-pressed glasses was examined using a stylus surface profilometer. The glass surface curvature was observed after spontaneous demolding by thermal shrinkage stress of the glass and the mold during cooling. The curvature decreased with increasing mold stiffness, as determined by the Young’s modulus and mold thickness. The curvature also decreased with decreasing cooling rate during cooling, which was discussed in terms of the glass fragility. It was demonstrated that the Young’s modulus, mold thickness, and cooling rate are key factors for the control of the molded glass surface flatness.     

一种快速的玻璃瓶口裂纹检测算法

基于机器视觉的图像处理广泛地应用于各个领域;为了能更快速、更准确地对玻璃瓶口进行无接触的检测,提出了一种快速的基于数字图像处理技术的玻璃瓶口裂纹检测算法;首先对获取图像进行预处理,然后根据迭代法获取最佳阈值,进行特征提取;接着引入矩形位移传感器,利用双圆周法寻找裂纹点,最后使用领域扫描确定裂纹;通过对大量玻璃瓶口裂纹的检测,证明算法在处理图像的速度上完全符合目前生产流水线的要求,而且不会漏检,在实际应用中获得了良好的效果.     

中国玻璃艺术的仿玉时代及其发展局限

汉代是中国玻璃发展史上的重要时期,在经济繁荣和政治稳定的条件下,中国的玻璃业得以存在和发展。汉代的玻璃制造以模仿为主。玻璃作为玉的替代品出现．其造型、质地无不显示出对玉的模仿。这不仅限制了玻璃的使用,更阻碍了玻璃艺术的发展。汉代玻璃艺术在原料配方、造型等方面均有所创新．扩大了玻璃制造业的规模．增加了产品种类。拓宽了玻璃的用途。     

Effect of scribe‐wheel dimensions on the cutting of AMLCD glass substrate

Abstract— The scribe‐and‐break method for glass cutting is widely used to separate individual liquid‐crystal‐display (LCD) panels from a larger motherglass substrate cell. Optimum glass‐scribing conditions including scribe‐wheel dimensions, scribing load, scribing speed, etc., have been determined based on practical manufacturing experience. However, there has been no systematic study to determine the scribing conditions necessary to avoid stray breakage. In this paper, the influence of the scribe‐wheel angle and diameter upon the scribing and breaking of an active‐matrix liquid‐crystal‐display (AMLCD) glass substrate, Corning Code1737F glass, was experimentally investigated for the case of simple single‐glass‐sheet separation. It was determined that an equation including the factors of scribe‐wheel tip angle and diameter can be used to predict median crack depth as a function of scribe load. It was further found that the breaking force of the scribed sheet was strongly influenced by the residual stress created during scribing. A wheel having a 130° tip angle and 4‐mm diameter demonstrated the best results for sheet separation in terms of the lowest breaking force without lateral crack propagation. In addition, increasing the time interval between scribing and separating was found to result in an increase in the breaking force required to separate the glass substrate.     

Mechanical properties and fracture analysis of glass substrate for PDPs

Abstract— A thermal shock test was carried out using high‐strain‐point glass substrates for plasma‐display panels. From the fracture analysis, the fracture stress was determined and compared with the initial edge strength. It was observed that the edge strength greatly decreased because of micro cracks that formed on the glass surface during testing. Therefore, it is important to prevent the formation of micro cracks in order to avoid failure as well as to minimize thermal stress in the process. Fatigue is also an important parameter in preventing the failure of the glass substrate when the stress on it lasts for a long period of time.     

A self-tuning microcontroller-based glass measuring instrument

This paper describes the development and implementation of a self-tuning microcontroller-based instrument for measuring different thicknesses of glass and glass substitutes. The instrument’s hardware was successfully tested in the laboratory on coated and uncoated glass under a variety of conditions. It is designed to provide high performance and high precision measuring capability to the system user. The obtained experimental data suggest that the proposed instrument surpasses existing techniques in its accuracy, reliability, simple operation, safety and measuring capability.     

Effect of geometrical characteristics of a scribing wheel on the bending strength of LCD glass substrates

Abstract— Scribing followed by a breaking method has often been applied to separate the individual liquid‐crystal‐display (LCD) glass panels from the mother cells. In some special occasions, especially for compact mobile devices, the bending strength of the glass substrates after breakage is the most important concern. In this study, the effect of the geometrical characteristics of the scribing wheel on the bending strength of LCD glass substrates was investigated. A volume‐crack effective coefficient was defined to estimate the effect of different geometrical characteristics of the scribing wheels. It was found that there is a strong correlation between the bending strength and the coefficient before chipping appears as the scribing load is increased. Thus, the coefficient can be used for the assessment of the geometrical effect of scribing wheels on the bending strength of glass substrates.     

平板玻璃实时智能自动化系统的应用

目前国内有关智能自动化技术的应用比较少。因此 ,介绍平板玻璃智能自动化系统的原理 ,并结合浮法玻璃的生产工艺进一步介绍了此系统在玻璃生产过程的具体应用 ,旨在为智能自动化技术在玻璃生产中推广应用而抛砖引玉     

基于分布式的玻璃缺陷检测技术研究及性能优化

玻璃检测速度的的提高会在短时间会产生大量图像数据,传统分布式框架MapReduce处理速度和及时性无法满足玻璃缺陷检测的要求。课题将MapReduce分布式框架运用到海量图像处理,设计阈值分割算法完成对玻璃缺陷图像的处理。通过添加数据划分模块使计算与存储本地化,加快数据处理的及时性。实验结果表明改进的MapReduce计算框架处理速度平均提高14.1%,能够对运行速度为600m/h的玻璃带进行在线检测,并检测出玻璃带上缺陷的个数、位置和缺陷的类型。