摘要
独立分量分析(independent component analysis,ICA)是基于信号高阶统计量的信号分析方法,它可以找到隐含在数据中的独立分量。在分析独立分量分析的基本模型及方法的基础上,讨论了有噪声信号的独立分量分析,使用最大似然估计对有噪声的ICA模型进行去噪处理,并研究了基于ICA的软门限图像去噪方法。在仿真实验中与其他的图像去噪方法进行了比较,突出了该方法在噪声方差较小时对非高斯信号的去噪优势。
Independent component analysis（ICA） is a signal analysis method based on signal＇s high order cumulants,it can find out the latent independent components in data. In this paper,we show how ICA can be used for image denoising. We model the noise-free image data by ICA, and denoise a noisy image by maximum likelihood estimation of the noisy version of the ICA model. This leads to the application of a soft-thresholding operator on the each independent component. Demonstration indicates that the proposed method gives better result compared to Wiener method.
出处
《信号处理》
CSCD
北大核心
2008年第3期381-385,共5页
Journal of Signal Processing
