基于小波变换的数字全息再现像质量提高方法  

A method for improving quality of digital holographic reconstructed image base on wavelet transform

作　　者：贾勤[1,2] 李志全[1] 刘雪飞[2] 

机构地区：[1]燕山大学电气工程学院,河北秦皇岛066004 [2]中国环境管理干部学院信息工程系,河北秦皇岛066004

出　　处：《燕山大学学报》2012年第5期409-412,427,共5页Journal of Yanshan University

基　　金：秦皇岛市科学技术研究与发展计划项目(201101A059)

摘　　要：数字全息再现像中存在的散斑噪声严重影响了数字全息的应用,通过分析边缘检测方法和小波变换阈值去噪方法的原理,提出了一种基于边缘检测的小波变换散斑噪声去除方法。首先利用高斯-拉普拉斯算法获得边缘图像,进而通过Neyman-Pearson准则获得自适应阈值,并采用改进折中阈值函数对边缘图像和非边缘图像小波系数进行处理,将两者处理后的小波系数相加,并进行反变换得到处理后的图像。研究结果表明,该方法能够较好地在去除散斑噪声的同时保留图像细节。Speckle noise in digital holographic reconstructed image affects the application of digital holography seriously.A speckle noise reduction method using wavelet transform base on image edge detecting is proposed by analyzing edge detection and wavelet threshold denoising in this paper.Firstly,the edge image is got by LoG algorithmand adaptive thresholds are got base on NeymanPearson criterion;then an improved compromise threshold function is adopted to process the wavelet coefficients of edge image and non-edge image,the processed wavelet coefficients corresponding to these images are added up;at last,the de-noised reconstructed image is got by inverse transformation.The results show that the method can reduce the speckle noise and keep the image detail well.

关 键 词：数字全息 散斑噪声 小波变换 折中阈值函数 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]