互相关DPIV技术及其在氧化沟模型中的应用

Cross-correlation DPIV and application in oxidation ditch model

重点研究二维DPIV技术的速度信息提取方法并进行算法实现。该技术基于灰度图像互相关法,在相邻时刻的序列图像中,灰度分布最相似的子窗口即是最佳匹配,粒子在像平面上的运动位移由两个子窗口中心距离确定。通过对互相关测度的优选既能兼顾粒子匹配成功率也能降低匹配算法的时间开销,通过优化算法的搜索空间极大地减少了互相关运算量。给出了标准图像仿真结果及其在氧化沟模型流速测试中的成功应用。实验结果表明,该算法适合于氧化沟模型流场及类似场合的测速。

The method for extracting velocity field is mainly investigated and the image-processing system of two-dimensional Digital Particle Image Velocimetry （DPIV） is developed. The algorithm is based on gray image cross-correlation and the most similar sub-windows on the two frames in serial gray images are reckoned as optimal matching. The displacement of the particle movement on image plane is thus determined by the distance of the two centers of sub-windows. On basis of large amount of experimental results, the normalized cross-correlation coefficient is selected as the cross-correlation measure, which greatly reduces the computational time and meanwhile holds the particle matching accuracy. Moreover. the search regions are optimized to reduce time consumption. The experimental results of standard images as well as the successful application in the actual experiment of oxidation ditch model prove that the researched algorithm of velocity extraction is suited for the measurement of the flow field such as oxidation ditch model.