【作者】 张工；

【导师】 梁桂华；

【作者基本信息】 大连理工大学 ， 船舶与海洋结构物设计制造， 2006， 硕士

【摘要】 PIV(Particle Image Velocimetry，简称PIV)技术作为一种非接触式流场测量技术，克服了以往接触式单点测试技术的局限性，能够进行瞬态平面二维流场的测试，是一种非常有发展前景的无扰动流场测量技术，借助于现代电子技术和计算技术的发展，近20年流动显示技术有了长足的进步。而作为PIV测试技术关键部分的图像后处理算法，决定了处理结果的准确性、快速性和可靠性，也是研究的热点问题。 PIV中的图像后处理算法主要是图像的匹配算法。而匹配算法又分为基于图像灰度的匹配算法和基于图像特征的匹配算法。PIV发展到今天，大多数的后处理软件都采用基于图像灰度的匹配算法，特别是其中的互相关算法，以其相对于其它算法对噪声不敏感、匹配位置准确、无方向二义性等优点，得到了广泛的应用。但是由于其互相关运算要进行全搜索图像匹配，它的运算量非常大，对图像进行匹配时耗费时间巨大，这对于PIV技术向实时化发展是很不利的。 本论文在传统的互相关算法的基础上，使用了一种改进的互相关算法，结合VC++的强大编程功能，编制PIV图像后处理软件，该软件的匹配算法在不必遍历整个搜索区的情况下就可以找到匹配点，省去了在除了匹配点之外的其他点上作的无用功，因而提高了计算速度，作了一种不必查遍整个搜索区就可以找到匹配区域的尝试，取得了比较满意的结果。本软件还采用了不同以往PIV图像后处理软件的错误矢量的去除方法，简化了算法，并且对错误矢量的修正效果明显。此外，本论文也对PIV图像后处理算法的未来发展做了一些设想。更多还原

【Abstract】 PIV (Particle Image Velocimetry, is called PIV), the technology to take one kind of non-contact flow field survey technology, has overcome the former contact single point test technology limitation, can carry on the transient state plane two-dimensional flow field test, is one kind extremely has the prospects for development free-stream flow field survey technology, with the aid of the modern electronic technology and the computation technology development, in the recent 20 years flow imaging technology make a considerable progress. As the PIV test technology essential part, the image post-processing algorithm, has decided the processing result accuracy, the rapidity and the reliability, also is the research hot topic.In the PIV image post-processing algorithm mainly is the image match algorithm. And image matches algorithm divide into based on the image gradation match algorithm and based on the image characteristic match algorithm. PIV develops today, the majority post-processing software all uses based on the image gradation match algorithm, specially correlation algorithm, is opposite by it in other algorithms to the noise insensitively, matches the position accurately, no ambiguity direction and so on, obtained the widespread application. But because its cross-correlation operation must carry on the entire search image, its operand is extremely big, carries on when the match to the image costs the time to be huge, this regarding the PIV technology to the solid current events development is very disadvantageous.The present paper in the traditional cross-correlation algorithm foundation, has used one kind of improvement cross-correlation algorithm, unifies VC++’s formidable programming function, write the PIV image post-processing software, this match algorithm does not need to carry on the entire field to find out the match spot, omitted the time of compute in non-match spot, thus enhanced the computation speed, made one kind of try that need not to look up the entire field of search to be possible to find matches region, has obtained the quite satisfactory result. This software also differed from the past the PIV image post-processing software in the method of improper vectors deletion and correction, simplified the algorithm, and was obvious to the wrong vector revision effect. In addition, the present paper has also made some tentative plans to the PIV image post-processing algorithm future development.更多还原