基于小波分解和分层FDR阈值的图像滤波算法  

On Image Filtering Algorithm Based on Wavelet Decomposition and Layered FDR Threshold

作　　者：谢雪晴[1] XIE Xue-qing(College of Information Engineering,Chongqing Industry Polytechnic College,Chongqing 401120,China)

机构地区：[1]重庆工业职业技术学院信息工程学院,重庆401120

出　　处：《西南师范大学学报（自然科学版）》2018年第3期97-101,共5页Journal of Southwest China Normal University(Natural Science Edition)

基　　金：重庆市教委科学技术研究项目(KJ1603701)

摘　　要：针对目前常用滤波算法对图像噪声滤除效果差的问题,提出了针对图像滤波的分层FDR阈值滤波算法.该算法对图像经小波分解后最底层的低频系数和各层水平、垂直、对角3个方向的高频系数分别进行多假设检验,并确定出各层各方向上的阈值,最后分别进行阈值化处理.通过对Lena,Cameraman图像进行对比仿真实验,结果表明该算法在信噪比、峰值信噪比、均方误差等方面均优于中值滤波、通用阈值算法、FDR算法和自由分布式FDR算法,具有较好的噪声滤除效果.For the current commonly used filter algorithm does not have good filtering effect to the image noise,a layered FDR threshold filtering algorithm aiming at the image filter has been put forward.This algorithm makes multiple hypothesis testing to the low frequency coefficient at the bottom layer after wavelet decomposition to the image,as well as the high frequency coefficients respectively in horizontal,vertical and diagonal directions of all layers,confirms the threshold value of each layer,and finally makes threshold processing respectively.Through simulation experiments to Lena and Cameraman images,the results show that the algorithm is better than the median filtering,general threshold algorithm,FDR algorithm and freely distributed FDR algorithm in the signal to noise ratio,peak signal to noise ratio,and mean square error,which has better noise filtering effect.

关 键 词：图像滤波 小波分解 分层FDR 阈值 假设检验 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]