组件针孔疵病检测仪及其算法研究

【作者】 杨敏；

【导师】 常本康；

【作者基本信息】 南京理工大学 ， 物理电子学， 2009， 硕士

【摘要】 三代像增强器是增强微弱光学图像的真空成像器件,是微光夜视技术的核心器件之一,应用领域广泛。由于GaAs外延材料生长及阴极组件研制过程中造成的缺陷,对三代微光像增强器的成像质量具有一定影响,所以优质的GaAs外延材料是实现激活高灵敏度阴极的先决条件,是提高像增强器成品率的基础。本课题所研究的组件针孔疵病检测仪可用于对GaAs基片和GaAs阴极组件表面瑕疵进行检测,统计不同区域的瑕疵信息,淘汰不合格产品,提高组件的性能。本文阐述了组件针孔疵病检测仪的硬件结构体系,说明了各个硬件的结构和功能;详细介绍了组件针孔疵病检测仪的软件设计模块,描述了各模块的作用,并实现软硬件交互。基于目标提取的原则,讨论了图像分割中阈值分割的基本理论和方法,并在此基础上提出一种可同时对GaAs基片和阴极组件表面瑕疵进行检测的算法;提出海量图片无缝拼接算法,利用坐标平移和图像缩放实现无缝拼接;为了克服单个瑕疵跨越多张图片导致的瑕疵个数误计,提出瑕疵整合算法,同时增加描述瑕疵信息的方式,从而准确而完整地获得GaAs基片或阴极组件表面的瑕疵信息。组件针孔疵病检测仪设计了友好的用户界面,并实现计算机自动控制,大大提高了检测效率和准确度,并节省了人力资源,有利于提高GaAs阴极组件的制备质量和效率,加快三代微光器件的研究进程。更多还原

【Abstract】 The third generation imaging intensifier is a widely used vacuum optical imaging device which is used to enhance the faint image .It is one of the core low-light-level night-vision devices. The defects which are caused in the growth of GaAs epitaxial material and the process of cathode component have a certain impact on image quality of the third generation imaging intensifier .So high-quality GaAs epitaxial material is the first requirement of achieving high sensitivity cathode and the basis of improving the yield of imaging intensifier. Component-Pinhole Blemish Detector (CPBD) is designed in this assignment for study. The blemishes on the surface of GaAs substrate and cathode component can be detected automatically by CPBD. The statistical data of blemishes in different regions can be given. CPBD eliminate substandard products and improve the performance of component.This dissertation expatiates on the hardware architecture of CPBD, explains the architecture and functions of each hardware and also introduces the software modules of CPBD in details, describes the role of each module, realizes the interaction of hardware and software. Based on the principle of object extraction, the basic theory and methods of threshold segmentation are discussed in the paper. A detection algorithm of GaAs substrate and cathode component is proposed on this basis. An image mosaic algorithm of large number of images is proposed. Seamless mosaic by coordinate translation and image resizing is realized. A blemish integration algorithm is proposed in order to avoid miss counting which ??aused by a single blemish that strides across some images .At the same time, some ways are added to describe the blemish. Blemish details of GaAs substrate and cathode component can be gained accuracy and completely.A Friendly interface is designed in CPBD. The realization of computer automatic control greatly improves the efficiency and accuracy of detection and saves human resources. This will promote the preparation quality and efficiency of GaAs cathode component and speed up the research process of the third generation low-light-level devices.更多还原