毕业设计基于图像的工件外形缺陷检测技术研究.doc

摘 要 随着图像处理技术的发展，缺陷检测在现代工业连续、大批量自动化生产中起着重要的作用，它涉及各种各样的工件检查、测量和分类等应用，如检测工件的表面是否有划痕、印刷品形体是否有缺陷、IC字符是否完整、电路板焊点是否完善、饮料瓶盖的画面是否完整等。传统的抽样检查已无法保证人们对产品高质量的要求，基于机器视觉的在线高速检测为其提供了可能。 本文首先介绍了缺陷检测技术的研究背景和发展历程，其次介绍了实现缺陷检测图像处理技术的算法，最后为了确定光盘表面划痕缺陷的存在，设计了一个基于数字图像处理的缺陷检测系统。首先采集图像，接着对采集所得的图像在Matlab仿真环境下进行灰度转换、去噪、增强和分割处理，最后在Matlab环境下计算处理后的缺陷图像的面积和周长，进而求得其圆形度来确定缺陷的存在，并通过圆形度的取值范围判定光盘的缺陷类型为划痕。 关键词：机器视觉，缺陷检测，表面划痕，圆形度 ABSTRACT With the development of image process technology,disfigurement test is playing a more and more important role in the continuous and large batch automatic prodution of modem industry. It involves all kinds of workpiece test,measure, classification and any other application such as testing whether there are nicks on the surface of workpiece,whether there are disfigurement on the presswork physique,whether the IC byte is integrited, whether the welding spot on the circuit board are perfect,whether the image of lid of a beverage bottle is integreted,etc.Traditional sample test can’t meet people’s requirement towards high quality of products,and online telling test based on machine vision makes it possible. The text,at the beginning, introduces the study background and developing process of disfigurement test technology, then introduces the arithmetics of image process technology of testing, and finally designs a disfigurement test system based on digital image process technology to ensure the existance of nick disfigurement on the surface of CD.The design can be discribed as follows:firstly, collect the image;secondly, use the image in gray convert, removing noise, amplifying and splitting under the Matlab simulation environment,finally,compute the area and circumference of the processed disfigured image,so as to get the degree of circularity of the image to ensure the existance of disfigurement and judge whether the type of disfigurement of the CD is nick according the range dimension of the degree of circularity. KEY WORDS: machine vision, disfigurement test,