一种自适应各项异性高斯滤波方法  

An Adaptive Anisotropic Gaussian Filter for Noise Reduction

作　　者：王怀野[1] 张科[1] 李言俊[1] 

机构地区：[1]西北工业大学航天学院,西安710072

出　　处：《计算机工程与应用》2004年第10期18-19,共2页Computer Engineering and Applications

基　　金：高等学校博士学科点专项科研基金(编号:20020699014);"英才计划"资助项目

摘　　要：为了提高图像的信噪比且尽可能保留图像边缘信息,论文提出了一种自适应各项异性高斯滤波方法,该方法根据各象素位置的灰度梯度值决定在这一点是采用各项同性还是各项异性高斯滤波,如果采用各项异性滤波,则由该处的灰度梯度方向角自适应决定各项异性滤波器的长轴方向。仿真实验表明,论文提出的自适应各项异性滤波器在边缘保持方面优于各项同性高斯滤波器。Using gaussian filter for noise suppression,the noise is smoothed out ,and the edge is also distorted.The pa-per proposes an adaptive anisotropic gaussian filtering algorithm.At each position of the processing image,the algorithm selects a gaussian filter or anisotropic filter by the gradient threshold,if the anisotropic gaussian filter is adopted,the orientation of the filter's long axes is also calculated according to the gradiant orientation in the position.The result in-dicates that the algorithm has good character in noise reduction and edge preserving.

关 键 词：各项异性滤波 高斯滤波器 信噪比 高斯噪声 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]