车身部件三维视觉测量技术的研究
文献类型：学位论文
| 作者 | 胡彦强 |
| 学位类别 | 硕士 |
| 答辩日期 | 2017-05-24 |
| 授予单位 | 中国科学院沈阳自动化研究所 |
| 授予地点 | 沈阳 |
| 导师 | 马钺 |
| 关键词 | 计算机视觉 三维测量 白车身孔槽 亚像素 激光条纹中心 |
| 其他题名 | Research on 3D Vision Measurement Technology of Car Body Parts |
| 学位专业 | 控制工程 |
| 中文摘要 | 本文以华晨中华汽车公司的委托项目“白车身关键尺寸控制点的三坐标在线测量”为依托，针对白车身关键定位孔槽的三维坐标测量进行了实验研究，主要研究工作如下： （1）测量系统的设计与标定技术。分析双目立体视觉数学模型，选用基于HALCON标准靶标的标定方法，对搭建的视觉系统进行标定，以其中一组靶标图像建立约束关系，极线校正的误差为0.657个像素。 （2）目标边缘的亚像素定位技术。数学分析孔槽求解参数，提出了边缘亚像素定位与参数求解流程，其中，针对复合孔槽轮廓组成的复杂性，本文在Ramer算法基础上提出两步法分割策略，大大提高了分割效率。 （3）激光条纹中心提取技术。分析了几种经典的激光条纹中心定位方法，基于极值法的灰度重心算法具有速度快、精度较高等优点而被广泛运用，但是此算法易受噪声影响，所以本文在此算法基础上提出了基于梯度的ROI边缘搜索算法定位条纹中心点，并采用基于统计分布对条纹中心提取质量进行评估。 为了提高系统鲁棒性，本文引入权重完成对条纹中心点集以及孔槽弧段边缘点云的拟合，采用基于施密特正交化法的空间降维法解决孔槽中心位姿定位的问题，通过对白车身A柱冲压件的测量实验表明，本文中所提出的方法能够满足白车身关键点测量的精度要求，为白车身关键尺寸控制点的在线测量提供了理论依据。 |
| 英文摘要 | In this paper, the three-dimensional coordinate measurement of the key white-body locating hole is carried out, based on the Zhonghua’s commissioning project ‘Three Coordinates On-line Measurement for the White-body Key Measuring Points’.The main research is as follows: (1) Measurement and measurement system design and calibration technology. The calibration model of binocular stereo vision is analyzed. The calibration system based on HALCON standard is used to calibrate the built system. The constraint relation is established with a set of target images. The error of the line correction is 0.657 pixels. (2)Sub-pixel positioning technology for the target edge. Using the mathematical analysis to obtain the parameters of the key hole, the paper puts forward the process of locating sub-pixel and parameter solving. Aiming at the complexity of the composition of the composite hole, a two-step segmentation strategy is proposed，based on the Ramer algorithm, which greatly improves the segmentation efficiency. (3)The extraction technology of stripe center. This paper analyzes several kinds of classical methods for positioning the centers. The gray center of gravity algorithm based on extreme method is widely used in the advantages of fast speed and high precision. However, this algorithm is susceptible to noise, so a gradient ROI-edge searching algorithm is proposed to locate the strip center point , and the quality is evaluated based on the statistical distribution. In order to improve the robustness of the system, this paper introduces the weight to fitting stripe centers and the edges, and uses the method of spatial dimension reduction based on Schmitt's orthogonalization to solve the problem of pose center pose location. The experimental results show that the method proposed in this paper can meet the precision requirements of the critical point measurement of white body, and provides the theoretical basis for the on-line measurement of key points of white body. |
| 语种 | 中文 |
| 产权排序 | 1 |
| 源URL | [http://ir.sia.cn/handle/173321/20515]  |
| 专题 | 沈阳自动化研究所_智能检测与装备研究室 |
| 推荐引用方式 GB/T 7714 | 胡彦强. 车身部件三维视觉测量技术的研究[D]. 沈阳. 中国科学院沈阳自动化研究所. 2017. |
入库方式： OAI收割
来源：沈阳自动化研究所
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。