|本期目录/Table of Contents|

[1]曾艺辉△,高鸣.基于Bayesian估计的小波自适应阈值方法对图像进行去噪处理的研究*[J].生物医学工程研究,2018,04:410-413.
 ZENG Yihui,GAO Ming.Wavelet adaptive threshold method based on Bayesian estimation for image denoising[J].Journal of Biomedical Engineering Research,2018,04:410-413.
点击复制

基于Bayesian估计的小波自适应阈值方法对图像进行去噪处理的研究*(PDF)

《生物医学工程研究》[ISSN:1006-6977/CN:61-1281/TN]

期数:
2018年04期
页码:
410-413
栏目:
出版日期:
2018-12-25

文章信息/Info

Title:
Wavelet adaptive threshold method based on Bayesian estimation for image denoising
文章编号:
1672-6278 (2018)04-0410-04
作者:
曾艺辉1△高鸣2
1.武汉市第四医院放射科, 湖北 武汉 430033;2.漯河市中心医院影像科,河南 漯河 462000
Author(s):
ZENG Yihui1GAO Ming2
1.The Radiology Department of The Fourth Hospital of Wuhan,Wuhan 430033,China;2.The Imaging Department of the Central Hospital of Luohe,Luohe 462000,China
关键词:
小波理论信号噪声图像去噪核磁共振成像氧摄取分数
Keywords:
Wavelet theory Signal noiseImage denoising Nuclear magnetic resonance imagingOxygen extraction fraction
分类号:
R318
DOI:
10.19529/j.cnki.1672-6278.2018.04.07
文献标识码:
A
摘要:
利用Bayesian估计的小波自适应阈值方法对图像进行去噪处理。通过高斯滤波和小波变换的三种方法(传统的硬阈值、传统的软阈值去噪、基于Bayesian估计的自适应阈值去噪)分别同时对加不同标准差σ的Rician噪声信号进行消噪处理,对比验证高斯滤波和传统小波阈值去噪的优劣,以及新的Bayesian估计自适应阈值小波去噪在磁共振成像(magnetic resonance imaging,MRI)图像信号去噪方面的优越性。小波去噪后的信号信噪比比高斯滤波去噪后信号的信噪比高,且均方根误差要低。采用基于Bayesian估计的自适应阈值小波去噪方法比采用的高斯滤波保留了更多有用信号,优化后的氧摄取分数(oxygen extraction fraction,OEF)值有一定程度增大,使结果更接近正电子发射型计算机断层显像(positron emission computed tomography,PET)测量金标准。成功完成信号和噪声分离优化,将一种新的基于Baysian估计的自适应小波阈值去噪应用到了功能核磁共振成像的降噪分析上,取得了不错的效果。
Abstract:
To study the image denoising by using the wavelet adaptive threshold method estimated by Bayesian.Rician noise signal with different standard deviation σ was denoised at the same time,respectively by three methods of Gauss filter and wavelet transform(soft threshold and hard threshold-traditional denoising,adaptive threshold denoising based on Bayesian estimation).The traditional wavelet threshold filter and Gauss denoising quality were verified,as well as the new Bayesian to estimate the feasibility and superiority of the adaptive wavelet threshold denosing signals in MRI image.The signal-to-noise ratio(SNR)of wavelet after denoised was higher than that of Gauss filter,and the mean square error(RMS) was lower.More useful signals were obtained with adaptive threshold denoising method based on Bayesian estimation than that with wavelet denoising method.The optimized oxygen extraction fraction(OEF)value had increased and the result was closer to the gold standard of positron emission computed tomography(PET).The signal,noise separation and optimization are succssfully completed in the experiment.A new Baysian based adaptive wavelet threshold denoising is applied to the noise reduction analysis of fMRI with good results.

参考文献/References

-

备注/Memo

备注/Memo:
(收稿日期:2018-07-23) 湖北省卫生计生委科研项目(WJ2017M208)。△通信作者Email:Yhzeng5261@163.com
更新日期/Last Update: 2019-01-29