期刊导航

论文摘要

非整数步长的分数阶微分滤波器在图像增强中的应用

Non-integer Step of Fractional Order Differential in The Image Enhancement Application

作者:黄果(四川大学)

Author:huangguo()

收稿日期:2009-12-07          年卷(期)页码:2011,43(1):129-136

期刊名称:工程科学与技术

Journal Name:Advanced Engineering Sciences

关键字:图像增强;非整数步长;分数阶微分;拉格朗日多项式

Key words:image enhancement; non-integer step; fractional order differential; Lagrange polynomial

基金项目:国家自然科学基金

中文摘要

为了改善图像增强的效果,本文提出了一种新的图像增强方法。一方面,该方法继承了传统分数阶微分的优良特性,打破了传统G-L定义中取整数步长的思想,采用了非整数步长,从而提高了分数阶微分数值计算的精确性;另一方面,该方法利用基于线性加权的拉格朗日多项式的分段插值方法来确定非整数步长像素点的灰度值,这样可以在一定程度上能起到降噪的作用。因此,本文提出的方法在增强图像细节和抗噪方面取得了较好的平衡点。实验结果表明,在无噪声的情况下,该方法较传统的分数阶微分算子在图像增强方面效果更好;在有噪声的情况下,该方法具有一定的噪声免疫能力。

英文摘要

In order to improve the effect of image enhancement, A new method of image enhancement was presented in this paper. Firstly, this method in this paper inherited the fine characteristics of the traditional fractional order differential, and broke the idea to take an unit step in G-L definition to use non-integer step, which increased the accuracy of the fractional order differential numerical calculation. What’s more, the method utilized the linear weighting method based on Lagrange polynomial section interpolation to determine the gray values of the non-integer step pixel. To a certain extent, it could play a role in noise reduction. Therefore, the method proposed in this paper had made a better balance between enhancing the image details and denoising. The experimental results shew that in the noise-free case, the method proposed in this paper compared with the traditional fractional differential operator in the image enhancement was better; in a noise case, the method had a certain amount of noise immunity.

关闭

Copyright © 2020四川大学期刊社 版权所有.

地址:成都市一环路南一段24号

邮编:610065