定阈值Wiener滤波在子波域图像去噪中的应用  

Application of prethresholding Wiener filtering to wavelet domain image denoising

作　　者：鲁瑞华[1] 杨明[2] 

机构地区：[1]西南师范大学电子信息工程学院,重庆400715 [2]西南师范大学计算机与信息科学学院,重庆400715

出　　处：《系统工程与电子技术》2005年第10期1705-1707,共3页Systems Engineering and Electronics

基　　金：重庆市应用基础研究资助课题(2001-6968);西南师范大学发展基金资助课题(swnuf2004006)

摘　　要：针对图像处理中的去噪问题,分析了Wiener滤波中的统计误差,提出了定阈值Wiener滤波算法,通过定阈值对图像作预处理,提高了滤波器去噪性能。分别用标准Wiener滤波法、逆滤波法及定阈值Wiener滤波法对退化图像进行处理的结果表明,在进行图像恢复时,标准Wiener滤波效果比逆滤波好,定阈值Wiener滤波则优于标准Wiener滤波和逆滤波。与标准Wiener滤波相比,逆滤波降低了峰值信噪比,而定阈值Wiener滤波则提高了峰值信噪比。信噪比增大,表明定阈值Wiener滤波法对噪声的抑制能力增强。Aiming at denoising in image processing and analyzing the statistical error occurred in Wiener filtering this paper presents prethresholding Wiener filtering algorithm to improve filter performance in image denoising using prethresholding as a processing step. The results obtained by standard Wiener filtering, inverse filtering and prethresholding Wiener filtering in degenerated image processing show that conducting the image restoration, standard Wiener filtering is better than inverse filtering and prethresholding Wiener filtering is much better than both standard Wiener filtering and inverse filtering. Compared with standard Wiener filtering, inverse filtering reduces peak signal to noise ratio （PSNR） and prethresholding Wiener filtering improves PSNR. An increased PSNR shows that use prethresholding Wiener filtering raises capacity of the filter in noise reduction.

关 键 词：图像去噪 定阈值Wiener滤波算法 图像恢复 统计误差 峰值信噪比 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]