基于面结构光的三维轮廓测量技术研究

CORC > 沈阳自动化研究所 > 中国科学院沈阳自动化研究所 > 光电信息技术研究室

题名	基于面结构光的三维轮廓测量技术研究
作者	刘陈龙
学位类别	硕士
答辩日期	2015-05-26
授予单位	中国科学院沈阳自动化研究所
授予地点	中国科学院沈阳自动化研究所
导师	朱丹
关键词	三维轮廓测量三维重建相移算法解相位
其他题名	Research of the three-dimensional surface measurement based on structured light method
学位专业	模式识别与智能系统
中文摘要	近些年，随着经济的发展和科技的进步，基于面结构光方法的物体表面三维轮廓测量技术以其非接触、高精度等优点逐渐成为三维测量的主流方法。本文以三步相移法为基础，研究了三维轮廓测量的具体实现过程，探索了其中的实现细节和处理方法。本文主要包括四部分内容：系统原理及设计方案，三维轮廓测量的实现，关键算法的研究和系统测量精度分析。根据所选择的测量方法，首先确定了测量系统结构及测量过程；针对三角测量原理测量条件难以满足的情况，介绍了一种非理想条件下的系统结构及测量方法；同时介绍了系统中硬件的控制方法和软件算法的具体流程。针对测量过程中的具体问题进行了深入研究。针对图像中存在阴影区域的问题，本文提出一种分割方法，能够实现阴影区域的准确分割；本文用边缘检测的方法实现了对物体位置的定位，并由物体的位置信息完成图像边缘的裁剪；针对正弦条纹图像的特殊结构，选择了有针对性的旋滤波方法对图像滤波；针对相机和投影仪的标定问题，应用了一种联合标定方法，得到了相机和投影仪的参数矩阵。解相位算法是测量的关键，也是本文的研究重点。本文在分析了基于区域解相位算法实现过程的基础上，针对孤立区域解相位算法中存在解相错误的问题，提出一种改进方法，将像素质量图原理与区域解相方法相结合用于解相处理；实验结果表明，本文算法可显著提高区域解相位方法的解相精度。针对三维轮廓测量的测量精度问题，本文分析了诸多因素对测量结果的影响，并介绍了一些减小误差的方法。为验证算法的有效性和可靠性，本文在matlab开发环境下对系统和算法进行仿真测试与验证。实验结果表明本文方法能够稳定、可靠地实现对物体表面三维轮廓信息的测量，实际应用仍在研究中。
索取号	TH166/L73/2015
英文摘要	In recent years, with the development of economy and the progress of science and technology, structured light based three-dimensional(3D) surface measurement methods have become the mainstream because of the non-contact and high-precision and other advantages. In this paper, under the direction of the three-step phase shifting algorithm we did some research in 3D surface measurement area, explored the details and implementations of that method. The paper mainly includes the following four parts, respectively the principle of system and design scheme, the implementation of 3D surface measurement, the research of key algorithms, and the discussion of system precision. According to the measuring method selected, the structure of our measurement system and hardware devices had been chosen. For triangulation measurement method the measurement conditions are difficult to satisfy, to overcome that problem, this paper introduced a new system structure and measurement method. Meanwhile, the hardware control method and software algorithm had also been introduced. In view of the specific problems encountered in the process of measuring, this paper did some deep research. For the shadow areas problem in image, a shadow region segmentation method was proposed and used in the paper. With edge detection method to determine the object position, we could cut the image to reduce the data size. Considering the special structure of stripe image, a special filtering method was selected for image processing. In order to calculate the parameter matrix of camera and projector, a calibration method was introduced. Phase unwrapping algorithm is the key to the surface measurement, and it’s also the key point of this paper. Through analyzing the process of region based phase unwrapping algorithm, we proposed a novel phase-unwrapping method which based on region divided methods and quality guide techniques. Experimental results had demonstrated that the proposed algorithm can improve the precision of unwrapped data remarkably. For the precision of the measurement method, this paper represented some influencing factors and introduced some approaches to reduce error. In order to verify the reliability of the proposed method, the existing samples were used to experiment in matlab software environment. The experimental results had shown that our method can get the surface data of objects stably and reliably. The practical application is in research.
语种	中文
产权排序	1
页码	63页
内容类型	学位论文
源URL	[http://ir.sia.ac.cn/handle/173321/16774]
专题	沈阳自动化研究所_光电信息技术研究室
推荐引用方式 GB/T 7714	刘陈龙. 基于面结构光的三维轮廓测量技术研究[D]. 中国科学院沈阳自动化研究所. 中国科学院沈阳自动化研究所. 2015.