基于局部协方差模型的小波图像降噪  

Wavelet image denoising based on the model of local covariance

作　　者：董湘君[1] 

机构地区：[1]广东技术师范学院自动化学院,广州510635

出　　处：《激光杂志》2010年第3期22-23,共2页Laser Journal

摘　　要：自然图像小波系数的统计特性表明,图像边缘具有很强的局部协方差结构。本文据此提出小波域的局部协方差模型,从噪声图像中估计原图像。实验结果表明,该方法无论从视觉效果上还是峰值信噪比指标上均明显优于维纳滤波及小波阈值法降噪。It indicates that there is a strong local covariance structure introduced by the edges from the statistical properties of the wavelet coefficients of natural images.We suggest model of local covariance in this paper and estimate the original image from the noisy image.The experimental results show that the method in this paper is obviously superior to the wiener filtering and the wavelet thresholding whether in visual effect or in PSNR.

关 键 词：局部协方差模型 小波降噪 维纳滤波 峰值信噪比 

分 类 号：TN248.1[电子电信—物理电子学]