Velocity Field Measurement by Flow Images with Fourier Transform

Abstract

The present paper describes the method of flow field measurement using two visualized images which are recorded in a small time difference. The typical four methods, the so-called path-line method, the spatio-temporal derivative method, the correlation method and the two dimensional FFT method, are compared from the practical point of view. Then, the theoretical consideration of FFT method was made in the case that the images are not freezed completely, which called the image data with nois here. It is found that this nois always gives a error velocity vector. The simple method for excluding the error vectors is proposed. The relationship between FFT method and the correlation method is also shown. The correlation function is calculated using FFT, and used it in the two dimensional Fourier Transform Method. The more exact correlation method taking accout of gray level data of images is shown.