白噪声干扰下复数图像快速NLM去噪算法仿真  

Simulation of Fast NLM Denoising Algorithm for Complex Images under White Noise

作　　者：蔡宇 CAI Yu(College of Mobile Telecommunications Chongqing University of Posts and.Telecom,Chongqing 401520,China)

机构地区：[1]重庆邮电大学移通学院,重庆401520

出　　处：《计算机仿真》2021年第10期388-391,418,共5页Computer Simulation

基　　金：重庆市教委科学技术研究项目(KJZD-K201802401)。

摘　　要：采用当前方法在白噪声干扰下对复数图像进行去噪处理时,存在去噪效率低和去噪效果差的问题。提出白噪声干扰下复数图像快速非局部均值(Non-local means,NLM)去噪算法,分析白噪声在复数图像中的分布规律,采用主成分分析法对复数图像进行降维处理,在预处理后的复数图像中获取样本区域,对其进行小波系数修正,通过调整图像频带内小波系数相似加权和,控制小波系数之间存在的相似度,与噪声标准差之间为正比关系,利用修正结果计算各小波系数在小波分解后高频子带内的相似度,根据计算得到的相似度调整小波系数,实现复数图像的快速去噪。仿真结果表明,所提方法的去噪效率高、去噪效果好。When the current method is used to remove the noise from the complex image with white noise interference,the denoising efficiency and denoising effect are not ideal.Therefore,a non-local means(NLM)denoising algorithm for complex images with white noise interference was put forward.Firstly,the distribution rule of white noise in the complex image was analyzed,and the principal component analysis was used to reduce the dimension of the complex image.After that,the region of samples in the preprocessed complex image was obtained.The wavelet coefficient was corrected.The similarity between wavelet coefficients was controlled by adjusting the similar weighted sum of wavelet coefficients in the frequency band,and it was proportional to the standard deviation of the noise.The similarity of wavelet coefficient in the high-frequency subband after wavelet decomposition was calculated by the modified results.According to the similarity,the wavelet coefficients were adjusted to realize the fast noise removal of a complex image.Simulation results show that the proposed method has high denoising efficiency and a good denoising effect.

关 键 词：白噪声 复数图像 去噪 小波系数 

分 类 号：P391[天文地球—地球物理学]