基于机器视觉的密封圈检测系统研发

发布时间：2017-12-26 19:11

本文关键词：基于机器视觉的密封圈检测系统研发　出处：《广东工业大学》2015年硕士论文　论文类型：学位论文

【摘要】：密封圈在生产的过程中会出现尺寸误差、表面缺陷等问题,随着自动化生产水平的提高和生产规模的扩大,采用传统的人工检测的方法已经远远无法满足生产需求。机器视觉具有非接触、高效率和高精度的特点,基于机器视觉的检测方法是一种全新的方法。随着科技的发展,采用机器视觉的方法来检测密封圈缺陷,能大大提高密封圈的检测效率与检测精度,降低企业的生产成本。因此,本文结合实际的应用,对基于机器视觉的密封圈检测系统进行了研究和开发,以实现对密封圈的缺陷检测。本文首先对机器视觉检测的理论和方法进行了介绍,并对当下密封圈检测技术的技术背景和发展状况进行了分析。接着根据密封圈实际的检测要求,结合Bresenham算法设计了一套使用于密封圈缺陷检测的算法,最后结合了软件和硬件开发了基于机器视觉的密封圈检测系统。本文研究的主要内容包括：对多种计算机的传统画线算法和Bresenham算法理论进行了分析,并做了一个简单的程序对比几种算法的效率,证明了Bresenham在计算机硬件的实现上具有明显的优势,它与其它算法对比优势就是不需要进行浮点运算。对图像二值化的理论进行了学习和了解,并深入研究了Otsu阈值分割算法,并采用该算法对本文的密封圈原图像进行了二值化的处理。尺寸检测方面,首先是对传统的尺寸检测算法作为了解、分析和对比,继而本文的算法采用的是基于Bresenham算法设计出来的一种尺寸检测方法。在缺陷检测方面,介绍了几种方法,并与本文设计的算法做对比,得出本文的算法具有效率高,误检率低的优势。在机器视觉系统的开发中,选了合适的光源、相机和镜头等硬件搭建了密封圈检测的实验平台；并使用VS2010作为软件系统的开发工具,用OpenCV作为图像处理的函数库编写了软件,并给出了软件的框架和处理的流程。通过实验的结果分析验证,本文所设计的基于机器视觉的密封圈检测系统具有较高的效率和精度,为密封圈企业的自动化检测提供了技术指导。
[Abstract]:The size and surface defects of seal ring will occur during the production process. With the improvement of automatic production level and the expansion of production scale, the traditional manual detection method is far from enough to meet production needs. Machine vision has the characteristics of non - contact, high efficiency and high precision. The detection method based on machine vision is a new method. With the development of science and technology, the method of machine vision to detect the defects of sealing rings can greatly improve the detection efficiency and accuracy of seals, and reduce the production cost of enterprises. Therefore, in this paper, a seal ring detection system based on machine vision is studied and developed in order to detect the defects of the seal ring based on the practical application. This paper first introduces the theory and method of machine vision detection, and analyzes the technical background and development situation of the present seal ring detection technology. Then, according to the actual detection requirements of the seal ring, combined with the Bresenham algorithm, we designed a set of algorithm used to detect the defect of the sealing ring. Finally, combined with the software and hardware, we developed a machine vision based sealing ring detection system. The main contents of this paper include: the traditional painting on a variety of computer line algorithm and Bresenham algorithm are analyzed, and the efficiency of a simple program comparison of several algorithms, it is proved that Bresenham has obvious advantages in the realization of the computer hardware, it is compared with other algorithms does not require floating-point advantage. The theory of image two valued is studied and understood, and the algorithm of Otsu threshold segmentation is deeply studied. And the algorithm is applied to the original image of the seal ring in two valued way. In terms of size detection, the traditional size detection algorithm is first understood, analyzed and contrasted. Then, this algorithm adopts a size detection method based on Bresenham algorithm. In the aspect of defect detection, several methods are introduced and compared with the algorithms designed in this paper. The results show that the algorithm has the advantages of high efficiency and low error detection rate. In the development of the machine vision system, choose the appropriate light source, camera lens and other hardware experimental platform is built to detect the sealing ring; and use VS2010 as software development tools, using OpenCV as the image processing library to write the software, and gives the framework and software process. The experimental results show that the machine vision based sealing ring detection system designed in this paper has high efficiency and accuracy, which provides technical guidance for the automatic inspection of the sealing ring enterprises.
【学位授予单位】：广东工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.41;TB42

【相似文献】