一种改进的小波阈值去噪算法及其仿真分析  

An Improved Wavelet Threshold Denoising Algorithm and Its Simulation Analysis

作　　者：李永军[1] 马立元[1] 段永刚[1] 

机构地区：[1]军械工程学院导弹工程系,河北石家庄050003

出　　处：《自动化仪表》2011年第7期21-24,共4页Process Automation Instrumentation

摘　　要：针对小波阈值去噪方法中传统软硬阈值法各自存在的缺点,提出了开方法阈值去噪法。该方法对阈值的处理采用开方逼近的思路,使其曲线能很快逼近硬阈值方法曲线,并具有较好的连续性,克服了硬阈值法不连续和软阈值法有恒定偏差的缺点。通过仿真试验将开方法与其他三种方法的去噪效果进行了对比。结果证明,新方法兼顾了软硬阈值法的优点,去噪效果更好,信噪比更高。Aiming at the disadvantages of traditional soft and hard threshold methods for wavelet threshold denoising,the new developed extracting threshold denoising method is proposed.By adopting extraction approach concept,this method makes the curve rapidly approaching the curve obtained by hard threshold method and offers better continuity;and overcomes the demerit of soft threshold method with constant deviation.Through simulation and experiments,the effects of extraction method and other three methods are compared.The result verifies that the new method possesses both merits of hard and soft threshold methods,offers better denoising effect and higher signal to noise ratio.

关 键 词：去噪 小波分析 软硬阈值法 折衷法 干扰信号 傅里叶变换 

分 类 号：TN911.4[电子电信—通信与信息系统]