【作者】 陈钊；

【导师】 郭永彩；

【作者基本信息】 重庆大学 ， 光学工程， 2006， 博士

【摘要】 DPIV(Digital Particle Image Velocimetry)在现代流场测试中发挥着重要作用,是流场测速研究的主要发展方向之一。DPIV具有全场、动态、非接触测量等优点,广泛应用于空气动力学、水动力学、燃烧传值以及流体机械等各种实验中。近十年来,随着光学技术、成像技术和计算机图像处理技术的迅速发展,常规二维DPIV日趋成熟。技术进步新的要求和流场测速新课题的挑战不断出现,对流场进行三维DPIV、多相DPIV和微DPIV测量的需求日益增强。国外在DPIV研究领域起步早、技术较为先进。国内研究总体上不及国外,成果转化相对滞后,但研究对象涉及了几乎所有的DPIV课题。继续深化、完善DPIV不仅具有重要的学术意义,从实践角度来讲,进行DPIV研究和系统实现对推动国内DPIV技术开发应用和产业化进程具有重要的现实意义和应用价值。论文围绕着DPIV方法及与其相关技术来展开,全面、系统地研究DPIV的理论基础及其实现方法,重点研究DPIV硬件子系统的参数选择及其相互间的匹配、基于灰度图像互相关二维DPIV速度信息提取方法、DPIV亚像素定位方法和2D-3cPIV方法及其实现中的若干关键技术。DPIV原理比较简单,但在应用DPIV和系统开发时,对系统硬件参数选择及其相互间的匹配有着极其严格的要求。DPIV实现的第一个步骤是图像采集,应考虑的主要因素包括:示踪粒子的直径大小、密度、形状、光散射性能、播撒均匀性及浓度等参数,相机的曝光时间、帧转移时间、分辨率、像元尺度、敏感光谱范围、快门时间等参数,镜头畸变系数、焦距和光学分辨率、景深,激光光源的脉冲间隔时间和功率等参数以及片光和体积光的形成光路。确定这些参数的原则是粒子对流动介质具有较好的跟随性,保证获得高质量粒子图像,有利于后续算法的处理。论文系统地论述了这些因素及其如何影响测量结果,总结了几个经验公式以利于这些参数的估算和设置。粒子图像匹配是DPIV速度信息提取过程的核心。论文重点研究了其中的匹配测度、提速算法和匹配策略。通过分析大量的实验结果,选用标准化相关系数作为图像匹配算法的相关测度,既能兼顾粒子匹配成功率也能降低算法的时间开销,通过优化算法的搜索空间极大地减少了互相关运算量。通过DPIV误差源的分析表明,对图像互相关峰值位置和标定板上目标像中心的亚像素定位可以提高测量精度。经典的亚像素定位方法基于高斯模型,这种拟合法的精度完全取决于粗匹配的相关测度分布。当拟合曲面较为平缓时,需要增大拟合窗口,测量精度可能下降。最小二乘匹配法同时考虑了灰度图像的灰度更多还原

【Abstract】 Digital Particle Image Velocimetry (DPIV) plays an important role in modern flow research and becomes one of the tendencies in modern fluid velocity survey. With main characteristics of whole field, non-intrusive and dynamic measurement, DPIV has widely used in such varied experiments as aerodynamics, hydrodynamics, combustion and hydromechanism. With the rapid progresses in optics, imaging and computer image processing technologies in a recent decade, conventional two-dimensional DPIV is becoming mature technology. The investigations of Multiphase, micro and three-dimensional DPIV are in great demands because of higher challenges of the new tasks. Foreign researches on DPIV started earlier and their technologies are therefore advanced. However, including almost all of DPIV aspects, domestic researches and their translations to applications lag behind. Hence, the study on DPIV not only has an important science values and a great economic values but also can impel the home applications and industrializations of DPIV.This thesis is mainly concerned with the technology of DPIV and the relative techniques. The researches on the theory and realization methods of DPIV are fully carried out. Especially, emphasis is put on such items as constructing of hardware elements and their influences on measurement result, methods of velocity extraction based on gray image, sub-pixel matching method and the technology of 2D-3cPIV.Though the principle of DPIV is simple, the selection and optimization of system parameters are stricter. Image acquisition is the first step to record the particle images and the main factors to be considered include such parameters as tracer particle diameter, density, shape, scattering performance, seeding uniformity and concentration; exposal time, frame transfer time, resolution, image element scale, spectrum, shutter time, lens distortion, focus length, optical resolution, depth of field of camera; laser pulse interval and power and the generations of light sheet and volumetric light source. The rules of selecting these parameters are a good tracking capability, high quality image and in favor of post-processing algorithm. These factors and their affection on the measurement result are discussed in detail. Furthermore, several formulae are summarized to estimate and set up these parameters.The particle image matching is crucial for determining the two-dimensional velocity field. Therefore this thesis focuses on the matching measures and matching更多还原