Speckle noise reduction in digital speckle pattern interferometry using Riesz wavelets transform

Abstract

In this work, we present an effective method for speckle noise reduction in digital speckle pattern interferometry (DSPI), which is based on a Riesz wavelet transform thresholding technique. Riesz wavelet transform is a steerable pyramid wavelet transform. Before Riesz-wavelet decomposition is applied to the noised image; the given coefficients undergo to thresholding technique, where appropriate threshold limit at each level and threshold method (hard or soft thresholding) are used to remove the noise; therefore, the denoised image is obtained by reconstructing thresholded Riesz wavelets coefficients. The performance of the denoising method is analyzed by using computer-simulated correlation fringes, and the results are compared with those produced by discrete wavelet transform thresholding technique. An application of the proposed method to reduce speckle noise in experimental data is also presented.