Wavelet Based Image Denoising with Mixed Laplace Model

The performance of various estimators, such as maximum a posteriori (MAP) is strongly dependent on correctness of the proposed model of noise-free data distribution. Therefore, the selection of a proper model for wavelet coefficients distribution is very important in the wavelet based image denoising. This paper presents a new image denoising algorithm based on the modeling of wavelet coefficients in each subband with a mixture of Laplace random variables (rvs). Indeed, we design a MAP estimator, which relies on the mixture distributions. Using this relatively new statistical model we are able to better capture the heavy-tailed nature of wavelet coefficients. The simulation results show that our proposed method yields better performance than several published methods visually and quantitatively.

Keywords:

MAP estimator , mixture model , wavelet transform