崔公哲 張朝霞 楊玲珍 王娟芬
摘 ?要: 為了有效濾除信號中的噪聲,在提出的軟、硬閾值函數(shù)去噪方法的基礎上,結(jié)合已有的改進小波閾值去噪算法,新構(gòu)造一種小波閾值函數(shù)。文中新構(gòu)造的閾值函數(shù)結(jié)合軟、硬閾值函數(shù)的優(yōu)點,有較好的降噪效果和靈活性。通過Matlab仿真對比幾種算法的信噪比(SNR)和均方誤差(MSE)去噪指標,結(jié)果表明,新構(gòu)造閾值函數(shù)的去噪指標優(yōu)于傳統(tǒng)閾值函數(shù),具有一定的實用價值。
關鍵詞: 小波變換; 閾值去噪; 噪聲濾除; 去噪指標; 閾值函數(shù); 仿真分析
中圖分類號: TN919?34; TP391 ? ? ? ? ? ? ? ? ? ? 文獻標識碼: A ? ? ? ? ? ? ? ? ? ?文章編號: 1004?373X(2019)19?0050?04
Abstract: In order to effectively eliminate the noise in the signal, a wavelet threshold function is constructed on the basis of soft and hard threshold function denoising methods, and in combination with the existing improved wavelet threshold denoising algorithm. The threshold function proposed in this paper combines the advantages of the soft and hard threshold function, and has better denoising effect and flexibility. The signal?to?noise ratio (SNR) and mean square error (MSE) denoising indexes of several algorithms were compared in Matlab simulation experiments. The experimental results show that the denoising indexes of the newly?constructed threshold function are superior to those of the traditional threshold function, and have a certain practical value.
Keywords: wavelet transform; threshold de?noising; noise filtering; denoising index; threshold function; simulation analysis
由于外界環(huán)境的干擾,導致在實際信號的采集過程中無法避免地引入一些隨機噪聲,從而影響下一步的信號處理,所以如何對含噪信號進行去噪處理,提取出對研究有用的信號,成為信號領域的一個重要研究課題[1]。小波變換在信號處理方面有很廣泛的應用。許多科研工作者對這種方法在信號處理方面進行了深入的探討。目前常用的對小波處理的方法主要有三種:模極大值重構(gòu)去噪法[2]、空域相關去噪法[3]和小波閾值去噪法[4]。其中,小波閾值去噪算法是應用最為廣泛的算法,也是被學者研究最多的算法。
在利用小波閾值函數(shù)去噪時,傳統(tǒng)的閾值函數(shù)主要有硬閾值函數(shù)和軟閾值函數(shù),軟、硬閾值法是由Donoho和Johnstone等人于1995年在小波變換的基礎上提出的,由于這兩種方法繼承了小波分析的優(yōu)點,而且計算量小,實現(xiàn)方法簡單,目前已得到廣泛的應用[5]。但是軟、硬閾值函數(shù)去噪算法由于本身的函數(shù)缺陷致使在信號去噪時并不能得到理想的效果。在小波閾值算法中選擇合適的閾值函數(shù)是小波閾值去噪算法需要解決的關鍵問題之一[6]。
本文針對軟、硬閾值函數(shù)的缺點,在文獻[8]的基礎上構(gòu)造出一種新的改進閾值函數(shù),將改進后的閾值函數(shù)和軟、硬閾值函數(shù)以及文獻[8]的閾值函數(shù)進行對比,證明本文提出的閾值函數(shù)具有較好的性能,通過仿真實驗證明了新閾值函數(shù)的有效性和優(yōu)良性。
從表1可以明顯看出,經(jīng)過新的改進閾值函數(shù)去噪后的Heavy sine曲線的信噪比(SNR)最大且均方差(MSE)是最小的。
本文在軟、硬閾值函數(shù)以及文獻[8]的基礎上構(gòu)造一種新的閾值函數(shù)來提高含噪聲信號的去噪效果。新的閾值函數(shù)克服了軟、硬閾值連續(xù)性差和存在固有偏差的缺點,同時由于[α]的存在使得改進閾值函數(shù)更便于調(diào)節(jié),靈活性好。利用上述幾種小波閾值去噪方法對Heavy sine信號進行仿真實驗,仿真結(jié)果表明,利用本文新構(gòu)造的函數(shù)進行去噪,可以很好地限制偽吉布斯現(xiàn)象,在信噪比(SNR)和均方誤差(MSE)性能指標上均優(yōu)于軟、硬閾值函數(shù),能夠較好地保留有用信號。
參考文獻
[1] 聶林濤,張愷.基于小波分析的一維含噪信號的除噪方法[J].兵器裝備工程學報,2011,32(12):86?88.
NIE Lintao, ZHANG Kai. Detection of one?dimensional noisy signal based on wavelet analysis [J]. Journal of ordnance equipment engineering, 2011, 32(12): 86?88.
[2] MALLAT S, HWANG W L. Singularity detection and processing with wavelets [J]. IEEE transactions on information theory, 1992, 38(2): 617?643.
[3] XU Y S, WEAVER J B, HEALY D M. Wavelet transform domain filters: a spatially selective noise filtration technique [J]. IEEE transactions on image processing, 1994, 3(6): 747?757.
[4] DONOHO D L. De?noising by soft?thresholding [J]. IEEE tran?sactions on information theory, 1995, 41(3): 613?627.
[5] 朱偉華,安偉,尤麗華,等.基于改進閾值函數(shù)的小波閾值去噪算法[J].計算機系統(tǒng)應用,2016,25(6):191?195.
ZHU Weihua, AN Wei, YOU Lihua, et al. Wavelet threshold denoising algorithm based on improved wavelet threshold function [J]. Computer systems & applications, 2016, 25(6): 191?195.
[6] 王蓓,張根耀,李智,等.基于新閾值函數(shù)的小波閾值去噪算法[J].計算機應用,2014,34(5):1499?1502.
WANG Bei, ZHANG Genyao, LI Zhi, et al. Wavelet threshold denoising algorithm based on new threshold function [J]. Journal of computer applications, 2014, 34(5): 1499?1502.
[7] ATTO A M, PASTOR D, MERCIER G. Wavelet shrinkage: unification of basic thresholding functions and thresholds [J]. Signal, image and video processing, 2011, 5(1): 11?28.
[8] 陳莉明.基于改進小波閾值算法的信號去噪[J].激光雜志,2015,36(3):92?95.
CHEN Liming. De?noising algorithm of signal based on improved wavelet threshold [J]. Laser journal, 2015, 36(3): 92?95.
[9] 張振鳳,威歡,譚博文.一種改進的小波閾值去噪方法[J].光通信研究,2018(2):75?78.
ZHANG Zhenfeng, WEI Huan, TAN Bowen. An improved wavelet threshold denoisingmethod [J]. Study on optical communications, 2018(2): 75?78.
[10] 李永軍,馬立元,段永剛.一種改進的小波閾值去噪算法及其仿真分析[J].自動化儀表,2011,32(7):21?24.
LI Yongjun, MA Liyuan, DUAN Yonggang. An improved wavelet threshold denoising algorithm and its simulation analysis [J]. Process automation instrumentation, 2011, 32(7): 21?24.
[11] 趙瑞珍,宋國鄉(xiāng),王紅.小波系數(shù)閾值估計的改進模型[J].西北工業(yè)大學學報(自然科學版),2001,19(4):625?628.
ZHAO Ruizhen, SONG Guoxiang, WANG Hong. Improved model of wavelet coefficient threshold estimation [J]. Journal of Northwestern Polytechnical University (Natural science edition), 2001, 19(4): 625?628.
[12] 葉裕雷,戴文戰(zhàn).一種基于新閾值函數(shù)的小波信號去噪方法[J].計算機應用,2006,26(7):1617?1619.
YE Yulei, DAI Wenzhan. Signal de?noising in wavelet based on new threshold function [J]. Journal of computer applications, 2006, 26(7): 1617?1619.
[13] 陶紅艷,秦華峰,余成波.基于改進閾值函數(shù)的小波域去噪算法的研究[J].壓電與聲光,2008,30(1):93?95.
TAO Hongyan, QIN Huafeng, YU Chengbo. Reseach of signal denoising method based on an improved wavelet threaholding [J]. Piezoelectrics & acoustooptics, 2008, 30(1): 93?95.