基于平滑阈值函数的小波图像去噪

本文在分析软阈值函数的基础上通过对软阈值函数进行修正，提出了平滑阈值函数。此种函数因在整个取值区间内保持平滑性，所以在去噪中可保存图像的部分细节。本文将此种函数与基于贝叶斯风险估计求阈值相结合，用于自适应的小波阈值去噪，提高了图像的信噪比，减少了图像的均方误差。试验结果证明了这一点。

Wavelet Image Denoising Based on the Smoothness Threshold Function

In this paper, after analysing the soft threshold function, the smoothness threshold function is proposed by modifying the soft threshold function. Because this function keeps smooth in the entire value range, some important details in the process of image denoising can be held. In this paper, we combine the function with the Bayesian risk estimator to find the threshold. It can be used for adaptive wavelet threshold denoising. Better MSEs and SNRs of image can be obtained according to the experimental results.