基于二阶段噪声检测的滤波算法研究  

Application of Filtering Algorithm Based on Two-stage Noise Detection

作　　者：赵国栋[1] 袁益民[1] 

机构地区：[1]上海电机学院,上海200240

出　　处：《计算机仿真》2011年第6期129-131,172,共4页Computer Simulation

摘　　要：研究信号优化问题,针对信号中的混合噪声,大大降低了信号的精度和准确性,影响了对真实信号特征的提取,为了消除存在于信号中的混合噪声,提出了一种基于二阶段噪声检测的滤波算法。先通过小波空域法处理由高斯噪声控制的小波系数;根据脉冲噪声控制的小波系数幅度的特点,结合中值滤波算法,构造一种能检测出脉冲噪声控制的小波系数的检测器,处理检测出的小波系数;最后用二次处理后的小波系数进行信号重构,可以去除混合噪声;仿真实验表明方法有效可行并优于单一的小波空域相关去噪法和软阈值去噪法,而且能够较好地去除混合噪声,达到提取信号优化过程。The mixed noise in signal reduces signal quality greatly and affect the signal feature extraction.In order to eliminate the noise,an effective spatial correlation denoising method is proposed.Firstly,applying the spatial correlation of wavelet denoising method,we process the wavelet coefficients controled by the Gaussian noise;secondly,under the characteristics of wavelet coefficients controlled by impulse noise and combining with median filtering algorithm,the author brings forward a detector detecting wavelet coefficients controlled by the impulse noise and deal with the wavelet coefficients.Finally,the denoising singal is reconstructed with the treated wavelet coefficients to achieve the purpose of removing mixed noise.Simulation results show that this method is effective and feasible and is superior to a single spatial correlation of wavelet denoising method and soft-threshold de-noising method,and is also better for removing the mixed noise and achieving the purpose of extracting the signal

关 键 词：脉冲噪声 空域相关滤波 中值滤波 软阈值 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]