1. 东南大学仪器科学与工程系,江苏,南京,210096
2. 南京邮电大学计算机学院,江苏,南京,210003
3. 东南大学仪器科学与工程系江苏南京,210096
4. 南京邮电大学计算机学院江苏南京,210003
纸质出版:2008
移动端阅览
周宁宁, 赵正旭, 秦文虎. 图像的中介滤波算法与图像中介保真度度量[J]. 电子学报, 2008,36(5):979-984.
ZHOU Ning-ning, ZHAO Zheng-xu, QIN Wen-hu. Medium Noise Removal Algorithm and Medium Measuring of Fidelity of Image[J]. Acta Electronica Sinica, 2008, 36(5): 979-984.
首先通过对图像噪声产生过程的分析
并根据随机性和模糊性的特征
提出新的图像噪声的定义;然后
以中介真值程度的数值化度量为基础
设计了新的图像去噪滤波算法.通过仿真实例
并采用经典的PSNR评价方法
与已有的典型算法处理结果比较表明
基于中介真值程度度量的图像去噪算法的去噪效果优越
因而更具有实用性和有效性.针对PNSR等经典评价方法的结果往往与人们主观视觉不一致的弱点
本文提出一种新的基于图像中介熵的图像保真度度量方法
对示例的评价结果表明
该方法符合人们的主观感受.
This paper makes a new definition on image noise according to two features of noise known as the random and fuzzy after the analysis of the progress of image noise production.Following that a new image noise removal algorithm based on the measuring of medium truth scale is given.By sample simulation and classical PSNR evaluation
it shows the image noise removal algorithm based on measuring of medium truth scale brings about a better noise removal effect compared with the classic one
which will lead to its practicable and effective applications in noise removal and image restoration.In addition
a new measuring method of fidelity of image based on image entropy is introduced in this paper to reduce the common deviation between the classical assessment results such as PNSR and the human subjective perception as well as its validity is proved by a sample assessment.
0
浏览量
1217
下载量
4
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621