基于光滑分割的局部脊波变换图像去噪算法研究  

Image de-noising algorithm based on local-ridgelet transform with smooth partition

作　　者：李京娜[1,2] 王国宏[1] 王刚[2] 

机构地区：[1]海军航空工程学院,山东烟台264001 [2]鲁东大学信息与电气工程学院,山东烟台264025

出　　处：《激光杂志》2012年第6期25-27,共3页Laser Journal

基　　金：国家自然科学基金项目(61102165;61102167);鲁东大学横向基金项目(2010HX007)

摘　　要：针对小波变换和(整体)脊波变换去噪性能的不足,本文着重研究基于光滑分割的局部脊波变换去噪算法,同时采用包含一个调整因子的改进阈值收缩法分别对图像进行去噪处理。结果显示,调整因子的取值对图像去噪效果有一定影响,根据图像结构和纹理特征,选取一个适当值,基于光滑分割的局部脊波变换能够在明显去除噪声的同时,较好地保留原始图像的特征信息。Image de- noising algorithm based on local- ridgelet transformwith smooth partitioning is researched in this paper to overcome the ddiciency ofwavelet transform arid ridgelettransform, which simultaneouslyemploys the threshold shrink de-noising algorithm in which an adjustment factor is introduced.The results demonstrate that the adjustmentfactor has a certain influenceto de - noising, andif its value is selected to be appropriateaccording to inage structureand texture features,local - ridgelet transform basedon smooth partition can excellentlyde - noise and preservecharacteristic infonnation of original image.

关 键 词：小波变换 脊波变换 局部脊波变换 阈值收缩 峰值信噪比 结构相似度 

分 类 号：TN248.1[电子电信—物理电子学]