An Automatic Imaging System for Evaluating the Particles on Microstructure Surface

• Main Authors: Wang, Chau Hong, 王昭宏
• Other Authors: Lin Chern Sheng
• Format: Others
• Language: zh-TW
• Published: 1996
• Online Access: http://ndltd.ncl.edu.tw/handle/27176400079446184529

碩士 === 逢甲大學 === 自動控制工程研究所 === 84 === The quality measurement on the microstructure surface of an device becomes more and more important during the process of high density fabricating factories, such as semiconductors, LCD, optical disc, and so on. But the inspections by human beings often waste time and energy, and the experimental results could not be analyzed and accounted efficiently. Similarly, the quantity, size, and shape measurements for the particles in clean room system are also important exercises. Therefore, in this paper, we applied computer vision and pattern recognition technique to solve the quality inspection problems. The inspection items which could be useful applications include: defect caution of an optical disk templete; quantity evaluation of the spacers on the LCD templete; size measurement and pattern analysis of particles in cleanroom system. In practice, an automatic imaging system is established for evaluating the physical quality of particles on a film or plate surface. We get the micro-particle gray images from a microscope and CCD camera. The x-y table supporting module and control system provide an automatic way for across-the-board scanning. Through a series of image processing, the properties of particles, such as area and perimeter measurement; quantity calculation; particle pattern recognition, were evaluated and discussed during our experiments. Finally, we applied the minimum distance classifier algorithm and the novel feature recognition neural network to solve the classification and recognition problems in particle images.