摘要
通过分析多普勒测速技术与互相关技术的区别,把互相关技术应用于尾流场气泡幕参数的测量研究中。用高速CCD相机拍摄实验模拟尾流图像,通过后处理计算气泡幕的速度信息,对常规的互相关算法进行改进,采用快速傅立叶变换(FFT)进行互相关函数的快速计算。同时,为提高计算精确度,对计算出的互相关最大值所在的位置进行亚像素拟合,获得了亚像素精度的气泡位移。为了验证改进的算法,对PIV-STD序列标准图像进行仿真计算和误差分析,使测得的气泡速度的平均误差小于±1%。结果表明,该方法是可行的,并可推广用于其它的流场测量。
Cross-correlation methods have been applied to the measurement of bubble film parameters in ship wake from a new aspect. The high-speed COD camera is used to acquire the images of simulating ship wake images, and the velocity of micro-bubble in the ship wake is computed from image posttreament. Through improving cross-correlation algorithm commonly used for analysis of flow field,the fast Fourier transform（FFT） has been used to fastly compute the cross-correlation function. The selection of sizing grid for computing cross-correlation is discussed, and curved surface fitting is applied to improve the computing level for the location of max value of cross-correlation function, and the motion direction of micro-bubble in ship wake flow with sub-pixel level is obtained.
出处
《光电子．激光》
EI
CAS
CSCD
北大核心
2005年第10期1231-1234,共4页
Journal of Optoelectronics·Laser
基金
国防科技重点实验室基金资助项目(51448030105ZK1801)
关键词
互相关
微小气泡
速度
尾流
亚像素
cross-correlation
micro-bubble
velocity
ship wake
sub-pixel
