基于贝叶斯估计的小波图像去噪算法研究

The Research of Wavelet Image Denoising Algorithm Based on Bayesian Estimation

利用拉普拉斯分布作为小波系数的分布模型,用贝叶斯最大后验概率估计法估计小波系数,给出了一种小波域自适应去噪算法.实验结果表明,本文算法与普通的软阈值法相比,可以获得更好的去噪效果.

Image denoising based on wavelet coefficients prior statistical model is hot point now, in this paper, an adaptive wavelet domain denoising algorithm based on Bayes estimate is proposed by exploiting Laplace distribution as the wavelet coefficients distribution model, and estimating wavelet coefficients by Bayes maximum a posteriori estimation（MPA）. Experiment results show that the algorithm has better denoising effect compared with ordinary soft threshold method.