基于改进谐波均值滤波的热中子图像降噪算法研究  

Research on Thermal Neutron Image Denoising Algorithm Based on Improved Harmonic Mean Filtering

作　　者：李帅 孟献才 洪兵 刘源煜 顾威 熊绪文 梁君筱 胡阳 LI Shuai;MENG Xiancai;HONG Bin;LlU Yuanyu;GU Wei;XIONG Xuwen;LIANG Junxiao;HU Yang(School of Computer Seience and Eingineering,Anhui University of Seienee and Technology,Huainan 232001,China;Institute of Energy,Heifei Comprehensive National Seience Center,Hefei 230001,China;Anhui Antian Lixin Engineering management Co.,LTD,Hefei 230001,China)

机构地区：[1]安徽理工大学计算机科学与工程学院,淮南232000 [2]合肥综合性国家科学中心能源研究院,合肥230031 [3]安徽安天利信工程管理股份有限公司,合肥230031

出　　处：《核电子学与探测技术》2024年第6期1008-1014,共7页Nuclear Electronics & Detection Technology

基　　金：合肥综合性国家科学中心能源研究院(安徽省能源实验室)项目(21KZS202、21KZS208);国家自然科学基金青年项目(12305200);安徽省住房城乡建设科学技术计划项目(2023-RK043)。

摘　　要：在热中子成像过程中,由于n/γ射线辐射、热噪声等因素的影响,导致中子图像产生大量噪声等降质现象,不利于无损检测和后续的定量分析。根据热中子成像的噪声特点,提出了一种改进的谐波均值滤波算法,用于热中子图像噪声去除。实验结果表明,该算法能有效去除热中子图像中的大尺寸白斑噪声,同时保留图像边缘信息,可为热中子成像图像的后续处理提供参考。In the process of thermal neutron imaging,due to the influence of neutron/gamma radiation,thermal noise and other factors,the neutron image inevitably produces large size visible white spot noise and other degradation phenomena,which is not conducive to non-destructive testing and subsequent quantitative analysis.According to the noise characteristics of thermal neutron imaging,an improved harmonic mean filtering algorithm is proposed for noise removal of thermal neutron images.The experimental results show that the algorithm can effectively remove large size white spot noise in thermal neutron image,while preserving image edge information,which can provide reference for the subsequent processing of thermal neutron image.

关 键 词：热中子成像 谐波均值滤波 盲图像质量评价 

分 类 号：T99[一般工业技术]