基于修正维纳滤波的小波包变换图像去噪

摘要

图像去噪是图像处理中一个非常重要的环节.为了改善降质图像质量,根据Donoho提出的小波阈值去噪算法,分析了维纳滤波原理,提出了一种基于修正维纳滤波的小波包变换图像去噪方法.利用修正维纳滤波对噪声图像进行处理,用处理后的图像计算噪声的标准方差,以此作为小波包的阈值.利用小波包对维纳滤波后的图像进行分解,实现对图像的低频和高频部分分别进行分解,用计算出的阈值对小波包树系数进行软阈值处理.利用小波包逆变换来获取去噪后的图像.结果表明:在噪声方差为0.01时,经该算法去噪后图像的PSNR比小波包自适应阈值去噪后的PSNR高出8.8 dB.该算法不仅能有效地去除加性高斯白噪声,而 且能很好地保留边缘信息,极大地改善了图像的视觉质量.%Image denoising is an important step in the field of image processing. In order to improve the quality of the degraded images, based on wavelet threshold denoising algorithm put forward by Donoho, the theory of Wiener filtering is analyzed and a denoising method using wavelet packet transforms based on the Wiener filtering is proposed. The noisy image is processed by the correctional Wiener filtering and the noise standard deviation is calculated by the remaining signal of Wiener filter to be regarded as the threshold of wavelet packet transforms. The image is decomposed into the low frequency part and high frequency part by using wavelet packet transform and the wavelet packet tree coefficients are processed with soft threshold by using the level dependent adaptive threshold. The denoising image is acquired by using wavelet packet inverse transform. The results indicate that, the Peak Signal-to-Noise Ratio (PSNR) gain of the proposed algorithm has reached 8.8 dB higher than denoising method on wavelet packet adaptive threshold when the noise variance is 0.01. The algorithm is more efficient in noise removal and edge reservation for all the noise images with different noise variances.