基于curvelet变换的红外图像去噪方法研究  

Research on Infrared Image Denoising Method Based on Curvelet Transform

作　　者：孙婷婷 崔少华[2] 孔令坤 董世稳 黄金乐 SUN Tingting;CUI Shaohua;KONG Lingkun;DONG Shiwen;HUANG Jinle(Department of Computer Science,Huaibei Vocational and Technical College,Huaibei 235000,Anhui,China;College of Physics and Electronic Information,Huaibei Normal University,Huaibei 235000,Anhui,China)

机构地区：[1]淮北职业技术学院计算机系,安徽淮北235000 [2]淮北师范大学物理与电子信息学院,安徽淮北235000

出　　处：《山西师范大学学报（自然科学版）》2023年第2期37-40,共4页Journal of Shanxi Normal University(Natural Science Edition)

基　　金：国家级大学生创新训练项目(202110373032,202110373036);安徽省自然科学研究项目(KJ2021A1376);安徽省质量工程项目(2020jxtd234,2020xskt171);安徽省人文社科重点研究项目(SK2020A0723)。

摘　　要：针对传统小波算法所生成的可分离小波只具有有限的方向,不利于图像去噪,本文采用curvelet算法对红外图像进行去噪处理.该方法以小波变换为基础对图像进行分解,在分解所得的一系列小波子带中,以定义的平滑窗函数对曲线边缘进行平滑分割,再对平滑分割处理的每个子块进行Ridgelet变换.最后将小波阈值范围外的系数置零,以curvelet逆变换对原始图像进行重构.红外图像的去噪实验表明,本文算法有效可行,相比传统的小波算法、SVD算法,本算法能获得更高的PSNR数值,去噪效果更佳.We proposed the curvelet algorithm in the field of infrared image denoising,to address the problem that the separable wavelet generated by the traditional wavelet algorithm only has limited direction,which is not conducive to image denoising.In proposed method,the image is decomposed based on wavelet transform.In a se⁃ries of wavelet sub⁃bands,the defined smoothing window function is used to smooth the edge of the curve,and then ridgelet transform is used to smooth each sub block.Finally,the band outside the wavelet threshold is set to zero,and the original image can be reconstructed by using the inverse curvelet transform.Compared with the traditional wavelet algorithm and SVD algorithm,this algorithm can get higher PSNR value and better denoising effect.

关 键 词：红外图像 去噪 CURVELET变换 分解与重构 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]