论文中文题名: | 大型工件特征点三维测量系统研究 |
姓名: | |
学号: | 06301 |
保密级别: | 公开 |
学科代码: | 081203 |
学科名称: | 计算机应用技术 |
学生类型: | 硕士 |
学位年度: | 2009 |
院系: | |
专业: | |
第一导师姓名: | |
论文外文题名: | Research on Targets Images Based Large-scale Work Piece 3D Measurement System |
论文中文关键词: | |
论文外文关键词: | Encoded target 3D vision measurement Edge extracting Target matching |
论文中文摘要: |
汽车的生产制造是一个超大规模的复杂制造过程,其最基本的前提是汽车零件的制造。传统的汽车零件设计制造流程要求有产品的蓝图设计或其CAD模型,然而在很多场合下设计的初始信息状态不是CAD模型,而是各种形式的物理模型或实物样件,若要进行仿制或再设计,必须对实物进行三维数字化处理并在CAD软件中对获得的数据建模。因此,准确、快速、完备地获取实物的三维几何数据,即对物体的三维几何面进行三维离散数字化处理,是实现汽车零件制造重要步骤之一。本文结合近年来新兴的图像序列三维测量技术,设计了一套大型工件特征点三维测量系统,用于实现三维几何面的三维数据测量。
首先通过对目前流行的三维测量技术的比较,确定了使用序列图像三维测量技术作为三维测量系统开发的理论基础。
其次对特征点三维测量系统进行了详细的需求分析及可行性分析。在设计过程中,本文将系统划分为三个模块:图像预处理模块,特征点匹配模块,三维重建模块。本文重点对图像预处理模块及特征点匹配模块进行探讨。介绍了相关的理论并详细论述了图像预处理模块中基于人工特征点的图像分割算法以及在特征点匹配模块中人工特征点匹配算法。对以上算法进行了实验验证。
最后,在上述研究成果的基础上,建立了大型工件特征点三维测量系统的整体框架,利用Matlab7.0开发了大型工件特征点三维测量系统。开发时留出接口供今后开发使用。运行结果表明,系统具有了三维测量的主要功能。
﹀
|
论文外文摘要: |
Automobile manufacture is a super-scale and complex process, which the basic premise is the production of cart-parts. The traditional design and product procedure requires the blue print or CAD model of part, however, the initial status of procedure is often without blue print or CAD model in many cases, instead there are physic models or original parts. In this situation, we must acquire the 3D data from the object and than modeling that in CAD software before the clone or designation. Therefore, acquiring 3D data precisely, rapidly and completely is the most important step in automobile manufacture. The paper analysis the technical of 3D measurement by encoded targets image sequence and experimentally testifies those technical by the 3D measurement system we developed.
Popular 3-dimension measure technologies are firstly compared, while the weaknesses are point out. Than 3D measurement by encoded targets image sequence is chosen as the theory base of 3D measurement system.
Secondly, the requirement and feasibility of the measure system have been done. In system designing, the system has been divided into three component modules: Image preprocessing module, feature point match module and space point reconstruction module. This thesis focuses on the first two modules, describes the methods of feature point extraction and feature point matching in detail.
Finally, based on the study and research above, we build the main frame of 3D measure system and give the class and interface in the software. Than we realize the system with Matlab7.0.The experiment result show the system possessed the basic function of 3D measurement.
﹀
|
中图分类号: | TP391.73 |
开放日期: | 2010-04-01 |