一种有效保留图像细节的自适应图像消噪方法

针对更多保留图像细节信息有效滤除噪声的问题，分析了双密度双树复小波的变换原理及特点，推导了双变量萎缩函数，提出一种基于双密度双树复小波变换的局域自适应图像消噪算法。首先对含噪图像进行双密度双树复小波分解；后根据小波系数的统计特性以及层内和层间系数的相关性，采用结合局域方差估计的双变量萎缩函数对小波系数进行处理，并用处理后的小波系数重构图像。实验结果表明：该算法在滤除噪声的同时可保留更多的图像细节，极大地改善了图像的视觉质量。

New adaptive method for image denoising with keeping details efficiently

In order to keep more details while denoising, an efficient local adaptive image denoising algorithm based on the Double-density Dual tree Complex Wavelet Transform (DD DT CWT) was proposed. The principles and characteristics of the DD DT CWT were analyzed and a Bivariate Shrinkage Function (BSF) was derived. The noisy image was firstly decomposed by the DD DT CWT, then according to the statistical properties of wavelet coefficients and the dependency of inter-scale with intra-scale coefficients, the wavelet coefficients were shrunk by the BSF with local variance estimation, and finally the denoised image was reconstructed by the shrunk coefficients. The experimental results prove that the proposed algorithm is efficient in noise removal and details reservation, and can improve the visual quality of the denoised image.