当前位置>主页 > 期刊在线 > 信息技术 >

信息技术2018年4期

基于Garrote 阈值法去噪的改进研究
刘春 ,安源 ,李欣
(大庆师范学院 计算机科学与信息技术学院,黑龙江 大庆 163712)

摘  要:针对传统Garrote 阈值函数采用固定阈值收缩高频细节系数,并对高频细节系数进一步收缩方面缺乏统一有效手段的问题,本文提出了一种基于传统Garrote 阈值法的改进去噪方法。该改进方法既能兼顾各尺度下的不同阈值,又能进一步收缩高频细节系数,并且易于实现、计算简单。在高斯白噪声去噪方面,去噪后的图像在均方差(MSE)和峰值信噪比(PSNR)上, 均优于传统Garrote 阈值法。


关键词:小波阈值去噪;阈值函数;均方差;峰值信噪比



中图分类号:TP751.1         文献标识码:A         文章编号:2096-4706(2018)04-0001-05


Research on the Improvement of Denoising Based on Garrote Threshold Method
LIU Chun,AN Yuan,LI Xin
(College of Computer Science and Information Technology,Daqing Normal University,Daqing 163712,China)

Abstract:Against the traditional Garrote threshold function which used fixed threshold to shrink high frequency detail coefficients,and which had the problem which lacked unified and effective method to shrink high frequency detail coefficients further, this paper proposed an improved de-nosing method based on traditional Garrote threshold method. This improved method can take into account the different thresholds at different scales,and can shrink high frequency detail coefficients further.And this method implemented easily,calculated simply. For the Gaussian white noise de-noising,the de-nosing image which used this paper’s method exceeded the traditional Garrote threshold at the in mean square(MSE)and peak signal to noise ratio(PSNR).

Keywords:wavelet threshold de-nosing;hreshold function;mean square error;peak signal


参考文献:

[1] David L. Donoho. Statistical estimation and optimal recovery [J]. The Annals of Statistics,1994,22:238-270.

[2] DONOHO DL. Asymptotic minimax risk for supnorm loss:solution via optimal recovery [J]. Probability Theory and Related Fields,1994,99(2):145-170.

[3] DONOHO DL. De-noising by soft-thresholding [J].Information Theory,IEEE Transactions on,1995,41(3):613-627.

[4] Ding Y,Selesnick I W. Artifact-free wavelet d e n o i s i n g:n o n - c o n v e x s p a r s e r e g u l a r i z a t i o n,c o n v e x optimization [J]. Signal Processing LettersIEEE 2015,22(9):1364-1368.

[5] 郭晓霞,杨慧中. 小波去噪中软硬阈值的一种改良折衷法 [J]. 智能系统学报,2008(3):222-225.

[6] 沙磊,任超. 改进的Garrote 阈值法去噪的研究 [J]. 工程勘察,2010,38(7):57-60.

[7] Breiman Leo. Better subset regression using the nonnegative garrote [J]. Technometrics,1995,37(4):373-384.

[8] DONOHO DL. Johm.tone IM. Ideal~patial adaptation via wavelet shnnkage. Biametnka,1994,81:425-455.

[9] MALLAT SG. A theory for multiresolution signal decomposition:the wavelet representation [J]. IEEE Trans Pattern Anal Mach Intell,1989,11(7):674-693.

[10] 戚国宾,甘雨,隋立芬,等. 利用小波阈值消噪改进M-W组合周跳探测性能 [J]. 测绘科学技术学报,2015,32(1):22-26+31.

[11] 谢杰成,张大力,徐文立. 小波图象去噪综述 [J]. 中国图象图形学报,2002,7(3):209-217.

[12] 屈中阳,李鸿光. 一种改进的集合平均经验模态分解去噪方法 [J]. 噪声与振动控制,2014,34(5):171-176.


作者简介:

刘春,教师,副教授,研究方向:数据库、图像处理。

安源,讲师,研究方向:软件工程、图像处理。

李欣,讲师,研究方向:软件工程、数据库。