A method of extracting the number of fringes/fringe spacing of interferogram based on wavelet matched filtering and Fourier transform technique is proposed. The edge images of the interference pattern of particle and the particle mask image are detected respectively by Mexican Hat wavelet and the center of particle can be extracted through the 2D correlation operation for the two edge images obtained. Then the interference pattern of each particle can be achieved using the center coordinate, shape and size of the particle image. The fringe counting/fringe spacing of the interferogram of a particle is obtained by Fourier transform and the modified Rife algorithm, and sub-pixel accuracy of the extracted frequency is acquired. The performance of the method is demonstrated using simulation and experimental measurements. The simulation results are that the recognition ratio is more than 90% and the extracted frequency error is less than 0.0185% for the overlap coefficient γ<11.62%. The measurement uncertainty is ±0.41 μm and the absolute error is 1.31 μm for the standard particles of diameter of 51.1 μm. The research results show that the algorithm presented in this paper is feasible.