摘要
在分数阶微分Günwald-Letnikov定义的基础上,设计了一种中心对称分数阶差分滤波器。同时考察了其对一维信号的滤波特性,进而给出了一种具有两个参数的用于图像边缘检测的滤波算子,并通过实验归纳了该滤波算子两个参数的作用。实验表明,在处理不含噪图像时,该方法能很好地提取图像边缘信息,其效果不亚于经典的边缘提取方法;对于含噪图像,由于具有两个可调的参数,此滤波器更为灵活,能够获得比其他一些经典方法更优的效果。
In this paper,based on Günwald-Letnikov fractional differential formula,a new filter with two parameters is designed for edge detection,which is called centre-symmetric fractional differential filter.First,analyze its amplitude-frequency characteristics and then introduce the functions of two variables by experiments on 2D images.At last,some conclusions are arrived at.samm In processing the images without noise,the method proposes performs as well as other classical edge extraction methods.Especially,it surpasses others when processing image with noise because of its flexible parameters.
出处
《微型机与应用》
2011年第9期46-49,53,共5页
Microcomputer & Its Applications
关键词
分数阶微分
中心对称
边缘检测
频幅特性
fractional differential
centre-symmetric
edge detection
amplitude-frequency
