广义全变分与巴特沃斯高通滤波融合图像去噪  

Image denoising method based on fusion of generalized total variation and Butterworth high pass filter

作　　者：乔闹生[1,2] 张奋[3] 

机构地区：[1]湖南文理学院物理与电子科学学院,湖南常德415000 [2]中南大学信息科学与工程学院,长沙410083 [3]湖南文理学院计算机科学与技术学院,湖南常德415000

出　　处：《计算机工程与应用》2014年第19期20-22,122,共4页Computer Engineering and Applications

基　　金：湖南省科技计划项目(No.2014FJ3027);湖南省教育厅科学研究重点项目(No.13A062);中国博士后科学基金(No.2012M521553);光电信息集成与光学制造技术湖南省重点实验室项目;湖南省"十一五"重点建设学科--光学基金;湖南文理学院重点建设学科--光学基金;湖南省光电信息技术校企联合人才培养基地资助项目

摘　　要：为了得到既可保留图像低频信息又可增强图像高频信息的去噪后的图像,提出了一种广义全变分范数与巴特沃斯高通滤波器融合的图像去噪方法。分析了广义全变分范数的图像去噪模型基本原理,给出了其优点与缺点;讨论了改进的巴特沃斯高通滤波器图像去噪方法,指出了其优点及不足之处;结合两者的优点,将两种去噪方法进行融合而得到了一种可增强对比度和保留边缘细节的图像去噪方法。计算机仿真的主观与客观实验结果证明了基本原理分析的正确性。In order to obtain the image after being denoised that can not only reserve the low frequency information of image but also enhance the high frequency information of image, an image denoising method based on the fusion of generalized total variation model and improved Butterworth high pass filter is proposed. Firstly, the basic principle of the generalized total variation model image denoising is analyzed, and its advantage and disadvantage are given. Secondly, the image denoising method of the improved butterworth high pass filte is discussed, and its advantage and disadvantage are also given. Finally, the two advantages are inosculated and an image denoising method that can not only enhance image contrast but also reserve image edge detail is obtained. Both the subjective result and the objective result of the computer simulation indicate the correctness of the basic principle analysis.

关 键 词：图像去噪 广义全变分范数 巴特沃斯高通滤波器 峰值信噪比 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]