机器视觉在笔记本电脑冲件检测中的应用-机械工程专业论文.docxVIP

II II 上海交通大学工程硕士学位论文 ABSTRACT THE RESEARCH OF MACHINE VISION TECHNOLOGY APPLIED IN NOTEBOOK METAL STAMPING INSPECTION ABSTRACT In modern industrial production, much inspection of their quality must be finished totally. So is the inspection about defect of Notebook metal stamping in this paper. In tradition, many of these activities were finished by people’s eye detection. Thus the cost of human and management was highly increased; and the demand of high-speed, on-line, highly precise inspection and “zero defects” could not be met. Therefore, study of machine vision on Notebook metal stamping has significantly practical value. This paper studied critical parts，light and the principle and type of aberrance in the machine vision detection. This paper emphasized on such image pretreatment as: grayscale image segmentation, getting and setting the value of threshold, image’s binarization, removing noise. Also several kinds of edge detection methods were strengthened and selected a best one applied in the project. In final according to the object, this paper researched edge detection mathematic model and method and selected the proper one： mathematical morphological edge detection method which fits in with inspection of Notebook metal stamping. A conclusion was drawn that taking advantage of reference object can calibrate quickly and efficiently rather than other methods; The mathematics model of edge detection and calibration method from this paper has an instructional meaning in similar product’s inspection and measurement. Keywords: Notebook metal stamping, Machine vision, System calibration, Edge detection 上海交通大学 上海交通大学 学位论文版权使用授权书 本学位论文作者完全了解学校有关保留、使用学位论文的规定，同意 学校保留并向国家有关部门或机构送交论文的复印件和电子版，允许论文 被查阅和借阅。本人授权上海交通大学可以将本学位论文的全部或部分内 容编入有关数据库进行检索，可以采用影印、缩印或扫描等复制手段保存 和汇编本学位论文。 保密已在一年解密后适用本授权书。 本学位论文属于 不保密面。 (请在以上方框内打 .J) 学位论文作者签名:彻问i 指导教师签名 z 传茶 日期: J.6lo 年 σ4月 4 日 日期 {o 年 .:6月。/日 上海交通大学 上海交通大学 学位论文原创性声明 本人郑重声明:所呈交的学位论文，是本人在导师的指导下，独 立进行研究工作所取得的成果。除文中已经注明引用的内容外，本论文 不包含