摘要
针对利用方向小波的多方向框架分割图像时形成的像素序列长度的不同,提出一种新的基于方向小波的差值滤波图像去噪算法。该算法根据白噪声分布规则,将像素序列分成两组,分别采用不同的阈值萎缩方法,并将所产生的方向子图像进一步的作差值滤波处理,最后对所有子图像进行线性平均。对含不同程度高斯白噪声的图像去噪仿真实验表明,与其他小波阈值去噪方法相比,该算法能更有效的去除噪声和保持图像纹理细节,信噪比提高1~3dB。
Image was partitioned into different length pixel sequences by using multi-directional frames. An algorithm based on directionalet and differential filtering for image denoising was proposed. According to Gaussian noise statistical distribution, the pixel sequences were divided into two groups and dealt with different threshoM method respectively. And then, the differential filter was used for all sub-images produced by directionalet and all filtered sub-images were averaged. Experimental results show that the proposed method performs better than other methods, visually and quantitatively.
出处
《系统仿真学报》
EI
CAS
CSCD
北大核心
2007年第9期2127-2129,2137,共4页
Journal of System Simulation
关键词
图像去噪
方向小波
多方向框架
多方向小波基
差值滤波
image denoising
directionalet
multi-directional frames
multi-directional bases
differential filtering