摘要
由于微观量子特性和随机性以及中子束注量率在时间和空间分布上都存在着一定的统计涨落,致使中子图像存在比较强的噪声.由于这种噪声的统计分布符合泊松-高斯混合模型,因此,提出了一种新的中子图像去噪方法.该方法结合了PPB加权最大似然估计算法与非线性方差稳定化变换,实现了中子图像的去噪复原,能够有效地抑制传统算法中的伪影现象并保证结果不失真.实验结果表明,该方法能够提供稳健的复原结果.
Because of the microscopic quantum properties and randomness,and the neutron beam fluence rate in time domain and spatial domain distribution,there is a certain statistical fluctuation,it produces strong noise.The statistical distribution accord with poisson-gaussian mixture noise model.There is a neutron image denoising method in this paper.This method combines the PPB weighted maximum likelihood estimation algorithm,and uses the nonlinear stabilizing of variance transformation,which realizes the neutron image denoising.This method can effectively inhibit artifacts compared with traditional algorithm and guarantee the fidelity of the result.The experimental results show that the method can provide robust recovery results.
作者
刘娜
乔双
孙佳宁
LIU Na;QIAO Shuang;SUN Jia-ning(School of Physics, Northeast Normal University, Changchun 130024, China;School of Mathematics & Statistics, Northeast Normal University, Changchun 130024, China)
出处
《东北师大学报(自然科学版)》
CAS
CSCD
北大核心
2018年第2期75-78,共4页
Journal of Northeast Normal University(Natural Science Edition)
基金
国家自然科学基金资助项目(11275046
11305034)
国家重大科学仪器设备专项基金资助项目(2013YQ04086101)
关键词
中子成像
泊松-高斯混合噪声
PPB加权最大似然估计
方差稳定化变换
neutron radiography
poisson-gaussian noise
PPB weighted maximum likelihood estimation
variance stabilizing transformation