PDTDFB变换和双变量模型在图像去噪中的应用

Image denoising using PDTDFB and bivariate model

对金字塔复方向滤波器组和贝叶斯最大后验估计理论架构下的双变量模型进行研究的基础上,结合二者的优点,提出一种新的图像去噪算法。PDTDFB(Pyramidal Dual-Tree Directional Filter Bank)变换具有近似时移不变性、多尺度、多方向选择性好的特点;双变量模型充分突出图像分解后系数的尺度内和尺度间的双重相关性;对噪声估计方法做出了详细阐述。仿真实验表明,与已有的多尺度理论(如:轮廓波等)和一些典型的图像去噪算法相比较,该算法的客观评价指标PSNR以及去噪后图像的主观视觉效果都有明显的提高和改善,能有效地保留原始图像的纹理和细节信息。

This paper proposes a new image denoising method based on the Pyramidal Dual-Tree Directional Filter Bank （PDTDFB） and the bivariate model under the framework of Bayesian MAP estimation theory.The proposed algorithm uses the PDTDFB＇s advantages of approximative translation-invariant,multi-scale and multidirection-selectivity,exploits the intra-scale and inter-scale correlations of PDTDFB coefficients, and elaborates the method of noise estimation.Compared with current multi-scale theo- ries（Contourlet, etc.） and some outstanding denoising methods, the simulation results and analysis show that the proposed algorithm obviously outperforms in both PSNR and visual quality,and effectively preserves detail and texture information of original images.