基于机器视觉的精确尺寸测量研究-机械电子工程专业论文.docxVIP

III III IV IV Abstract coordinate system. Finally, the world coordinates of the edge pixels are fitted to straight lines by means of linear fitting based on total least-squares, so as to acquire the geometry parameter of the plane workpiece. Through the analysis of experimental, the accuracy of the measurement can be analyzed through experiment. The result of the experiment indicates that this method has characteristics of strong anti-noise property, high detection precision and good measurement accuracy. 3、Through the analysis of sub-pixel edge detection algorithm, this paper presents an improved sub-pixel edge detection algorithm based on Gaussian surface fitting. The main object of this paper is the plane workpiece whose basic geometric elements are straight lines. The gradient amplitude distribution of the image edge in the direction of the gradient can be approximated as the Gaussian surface distribution, so the sub-pixel positions of the edge pixels can be obtained through the parameters estimation of the Gaussian surface. This paper introduces the theoretical basis of the algorithm, gives the derivation process and validated by Matlab simulation experiment. The algorithm is analyzed by the data of results, and compared with other methods. Key Word: camera calibration; edge detection; total least-squares; Gaussian surface fitting; sub-pixel 目 录 目 录 摘 要 I Abstract II 第一章 绪 论 1 §1.1 课题的研究背景及意义 1 §1.1.1 机器视觉技术 1 §1.1.2 摄像机标定技术 1 §1.1.3 边缘检测技术 2 §1.2 国内外研究现状 3 §1.2.1 机器视觉技术 3 §1.2.2 摄像机标定技术 4 §1.2.3 边缘检测技术 5 §1.3 本文主要研究内容与结构 5 §1.3.1 本文主要研究内容 5 §1.3.2 本文结构安排 6 第二章 摄像机标定 9 §2.1 摄像机标定的基本理论 9 §2.1.1 摄像机成像几何模型 9 §2.1.2 标定中常用的坐标系及相互关系 11 §2.1.3 透镜畸变及其模型 13 §2.2 摄像机标定方法的分类 14 §2.3 摄像机标定实验 14 §2.3.1 摄像机标定过程 14 §2.3.2 标定结果 15 §2.4 本章小结 16 第三章 基于边缘检测的视觉测量技术 17 §3.1 边缘检测概述 17 §3.1.1 边缘定义及类型 17 §3.1.2 梯度 19 §3.2 经典边缘检测算子 19 §3.3 边缘检测的一般步骤 21 §3.4 基于总体最小二乘拟合的工件几何参数测量方法 22 §3.4.1 总体最小二乘概述与原理 22 §3.4.2 工件几何参数测量的原理及过程 24 V V 目 录 §3.4.3 实验结果分析