小波阈值改进及在光谱信号去噪处理中的应用  

Improvement of Wavelet Threshold and Application on Spectrum Signal De-noising

作　　者：尚小燕[1] 韩军[1,2] 王松 孔英秀[4] 

机构地区：[1]西安工业大学光电工程学院,西安710032 [2]西安电子科技大学技术物理学院,西安710071 [3]中国兵器第二○八研究所,北京102202 [4]西安思源学院能源与电气工程学院,西安710038

出　　处：《科学技术与工程》2011年第1期42-45,共4页Science Technology and Engineering

摘　　要：针对宽光谱膜厚监控系统输出的光谱信号,通过改进Donoho阈值函数,对阈值加入微调因子,使与噪声幅度相近或小于噪声幅度的信号小波系数阈值减小,有利于其小波系数的保留。另一方面使噪声的小波系数的阈值增大,有利于其小波系数的滤除,使小波阈值滤波算法具有自适应性。通过实验表明该算法有效地抑制噪声,很好地保留了信号的细节信息,信号的峰值误差为0.7%—1.0%,峰位误差为0.1%—0.3%;提高了系统的监控准确度。Aiming to output spectrum signal of the thin-film thickness wideband monitoring system,Donoho threshold function is improved by adding micro-alignment factor to the threshold.It is beneficial for preserving wavelet coefficients of real signal whose amplitude is similar to or less than niose to make wavelet coefficient threshold of the signal decrease;On the other hand,wavelet coefficient threshold of noise increases for filtering out noise favorably.So wavelet threshold filtering algorithm has self-adaption.By experiment,the result shows that random noise is filtered well and the signal details are reserved perfectly,the peak error of signal is 0.7%—1.0%,the peak location one is 0.1%—0.3%,the system accuracy is improved.

关 键 词：Donoho阈值函数 微调因子 小波阈值滤波算法 自适应性 

分 类 号：TN25[电子电信—物理电子学]