基于机器视觉的空心圆柱零件同心度检测技术研究
文献类型：学位论文
| 作者 | 费菲 |
| 学位类别 | 硕士 |
| 答辩日期 | 2010 |
| 授予单位 | 中国科学院上海光学精密机械研究所 |
| 导师 | 何红 |
| 关键词 | 机器视觉 空心圆柱零件 同心度 图像处理 亚像素 |
| 其他题名 | Research on measurement technique of concentricity of Hollow-Cylindrical Parts based on Computer Vision |
| 中文摘要 | 机器视觉技术是一项新兴的“光机电一体化”技术，由于其高精度和高自动化的优点逐渐在检测行业等诸多工业行业中应用。轴套类零件的同心度检测是工业检测中的一项重要内容，国内零件生产企业的该项检测还处在人工操作或者用半自动和手工结合的阶段，需要大量人力参与，检测质量也难以保证。本项目中针对具有轴套类零件几何特征共性的空心圆柱零件，利用机器视觉技术，对零件的同心度及其内外圆尺寸进行了非接触检测方面的研究和实践。 本研究的主要工作内容包括： 1． 系统硬件方案的设计和搭建：分别从图像采集单元、光学系统单元和光源及照明单元进行考量设计，根据本项目待测零件离散性的特点和高精度的检测要求，对工业相机进行选型；根据零件的高厚度特征对拍摄图像像质进行分析，选择远心镜头作为成像设备，结合待测零件的尺寸和相机型号，估算视场大小和放大倍率，对镜头进行选型；比较正向和背向照明、准直和非准直光在拍摄图像上的区别，结合项目对精度的要求和需要检测的内容及价格等因素最后根据定下最终的照明方案——背向照明+同轴红光源。 2． 系统的软件编制：在Visual C++ 6.0 开发环境下编写零件检测的软件，其中根据拍摄到的图像特征，改进了传统图像处理流程，删减了改善像质的预处理过程，增加了边缘优化过程。 3． 对零件进行实物检测，检测结果达到亚像素级。 4． 对检测系统性能进行初步分析和研究：分析了系统检测精度，同时讨论了毛刺的尺寸、聚集度和数量、曝光时间和镜头畸变对结果精度的影响；分析了系统检测速度，根据零件处理的运行时间并结合其他部分处理响应时间，估算出系统的检测效率。 5． 传输和控制部分的设计：根据系统检测的方式——来料检测，提出了完成“来料—拍照—采集”信号控制的基本设计思路；结合照明方式提出采用透明传输介质的思路。 |
| 英文摘要 | Computer Vision is an emerging optical-electronic-machinery integratingtechnology. This technology is now has been applied in many parts of the industrial trades such as industrial inspection because of its high precision and automation. The concentricity measurement of Axlebox Parts is very important in industrial inspection. In many domestic Parts-Manufacturing enterprises, manual operation and semi-automation are still the main methods used on concentricity measurement of Axlebox Parts, which needs lots of labor and also can hardly guarantee the measurement veracity. Aiming at Thick Hollow-Cylindrical Parts which has the geometric commonness of Axlebox Parts, some research on non-contact measurement of concentricity and diameter dimension of Thick Hollow-Cylindrical Parts has been done in this paper, using the method based on Computer Vision Technology. The main work of this research includes 5 parts: 1. The choice of the System Scheme: choose the proper camera based on the discrete characteristics of checking samples and the high precision of measuring demand; choose the proper telecentric lens based on the thick characteristics of the samples and also the size of field and magnification calculated from the size of samples and the camera; choose the proper lighting scheme—a red coaxial light source with backward direction, which is based on the comparison results of forward and backward illumination mode, collimated and non-collimated lighting resources and also the measuring demand and price comparison. 2. Software Programming: complete an application program for the measurement of concentricity and diameter dimension of the samples based on Visual C++ 6.0; use an improved image-processing procedure according to the image of the samples, removing the pretreatment and adding a new procedure for edge optimization. 3. A measurement for mechanical parts has been done according to the design scheme. The measurement result showed that the precision can reach sub-pixel precision. 4. Analysis and study on performance of the measurement system: factors which may effect the precision of the measurement results, such as size of burrs, aggregation degree of burrs, quantity of burrs, exposure time and lens distortion; analyze the operating time and estimate the efficiency of the system. 5. The designing of the transmission and controlling part: present the controlling idea of “sample supplied-pictured-acquired” based on the incoming material measuring characteristics; present the idea of using transparent material for transferring base on the choice of backward illumination mode. |
| 语种 | 中文 |
| 源URL | [http://ir.siom.ac.cn/handle/181231/16672]  |
| 专题 | 上海光学精密机械研究所_学位论文 |
| 推荐引用方式 GB/T 7714 | 费菲. 基于机器视觉的空心圆柱零件同心度检测技术研究[D]. 中国科学院上海光学精密机械研究所. 2010. |
入库方式： OAI收割
来源：上海光学精密机械研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。