组合层内和层间小波系数的图像去噪方法

Image Denoising Based on Interscale and Intrascale Wavelet Coefficients

建立了能利用系数层间和层内相互依赖的关系模型,推导了应用最小均方误差原理去除图像中噪声的算法,而且运用隐参变量描述方式,使得小波系数那种反映图像边缘强度特征的空间变属性表现了出来,使算法具有空间自适应能力,实验结果显示该方法比目前存在的一些方法更具优越性。

This paper proposes an approach to exploit the strong wavelet interscale and intrascale dependencies,and deduces a denoising algorithm under the linear minimum mean square-error estimation scheme,moreover,it employs a hidden variable expression to adapt the spatially variational edge property of image wavelet coefficients.Finally,a spatial adaptive algorithm is implemented to remove the noised image.The experiment results show that the proposed approach outperforms some existing denoising schemes.