Abstract

Speckle noise reduction is a crucial pre-processing step for a successful interpretation of images corrupted by speckle noise, and thus, it has drawn a great deal of attention of researchers in the image processing community. The Bayesian estimation is a powerful signal estimation technique and has been widely used for speckle noise removal in images. In the Bayesian estimation based despeckling techniques, the choice of suitable signal and noise models and the development of a shrinkage function for estimation of the signal are the major concerns from the standpoint of the accuracy and computational complexity of the estimation.
In this thesis, a low-complexity wavelet-based Bayesian estimation technique for despeckling of images is developed. The main idea of the proposed technique is in establishing suitable statistical models for the wavelet coefficients of additively decomposed components, namely, the reflectance image and the signal-dependant noise, of the multiplicative degradation model of the noisy image and then in using these two statistical models to develop a shrinkage function with a low-complexity realization for the estimation of the wavelet coefficients of the noise-free image.
A study is undertaken to explore the effectiveness of using a two sided exponential distribution as a prior statistical model for the discrete wavelet transform (DWT) coefficients of the signal-dependant noise. This model, along with the Cauchy distribution, which is known to be a good model for the wavelet coefficients of the reflectance image, is used to develop a minimum mean square error (MMSE) Bayesian estimator for the DWT coefficients of the noise-free image. A low-cost realization of the shrinkage function resulting from the MMSE Bayesian estimation is proposed and its efficacy is verified from the standpoint of accuracy as well as computational cost.
The performance of the proposed despeckling scheme is evaluated on both synthetic and real SAR images in terms of the commonly used metrics, and the results are compared to that of some other state-of-the-art despeckling schemes available in the literature. The experimental results demonstrate the validity of the proposed despeckling scheme in providing a significant reduction in the speckle noise at a very low computational cost and simultaneously in preserving the image details.