基于贝叶斯估计的图像去噪算法研究

Image Denoising based on Bayesian Estimation Algorithm

本文提出了一种基于贝叶斯最大后验概率估计(Bayesian MAP)的图像去噪方法。通过Matlab软件仿真,对均值滤波、中值滤波、小波阈值去噪和本文提出的图像去噪方法进行分析比较。实验表明:本文提出的方法根据图像和噪声的特点,在小波变换之后,对其中的高频系数进行贝叶斯最大后验概率估计,比其他几种图像去噪方法更能提高去噪后图像的峰值信噪比,更好地保留了图像的细节特征,取得了较好的视觉效果。

This paper presents a Bayesian maximum a posteriori estimation （Bayesian MAP） method for image denoising.By Matlab software simulation, mean filtering, median filtering, wavelet thresholding denoising and image denoising method presented in this paper are analyzed and compared.Experiments show that：the proposed method is based on the image and noise characteris-tics, after the wavelet transform, Bayesian maximum a posteriori estimation is applied to the high -frequency coefficients, compa-ring with several other image denoising methods this paper can enhance the peak signal to noise ratio after denoising , and better re-tain details of the image features, and achieved good visual effect.