摘要
复数小波变换具有平移不变性和多方向选择性,适用于图像去噪。对自然图像复数小波系数统计分布进行建模,提出了一种新的概率密度函数。在此先验分布的基础上,通过运用最大后验概率估计方法从含噪系数中恢复原始系数,来达到滤出噪声的目的。数值实验表明本方法在去除噪声的同时保留了图像的细节信息,取得了很好的降噪性能,其峰值信噪比(PSNR)在高噪声水平下(n=30),较其他常见方法至少高1.9dB左右。
Complex wavelet transform has translation invariance and directional selectivity, and it is suitable for image denoising. The model based on statistical distribution for complex wavelet coefficients of nature images is set up and a new probability density function is proposed. Under such prior distribution, Maximum A Posteriori (MAP) estimator is used to restore the wavelet coefficients from the noisy observations. Numerical experiments show that the proposed method can remove the noise while preserving significant image details. At high noise level (σn=30), the method achieves at least 1.9dB gain over current leading methods for PSNR measurement.
出处
《光电工程》
CAS
CSCD
北大核心
2004年第8期69-72,共4页
Opto-Electronic Engineering
关键词
图像降噪
小波变换
统计模型
图像处理
Image denoising
Wavelet transform
Statistical model
Image processing
