基于分数阶微分的高斯噪声图像信息提取算法  

Research on Information Extraction Algorithm of Gaussian Noise Images Based on Fractional Order Differentiation

作　　者：侯明亮[1] 

机构地区：[1]淮海工学院计算机工程学院,江苏连云港222005

出　　处：《计算机与现代化》2010年第9期5-8,共4页Computer and Modernization

基　　金：江苏省高校自然科学基金资助项目(08KJD520013)

摘　　要：考虑分数阶微分对信号处理的性质与特点,提出一种基于分数阶微分的信息提取算法。首先针对各种系数高斯噪声的图像进行信息提取,并与Roberts、Prewitt、Sobel整数阶经典算子对图像信息提取的结果进行分析比较。实验证明,利用分数阶微分不但能提取整数阶算子的高频边缘信息,还能提取平滑区域的纹理信息,并对高斯噪声具有一定程度的抗噪性。Considering the trait of fractional order differentiation on signal processing,an information extraction algorithm based on fractional-order differentiation is put forward.To start with,by this means information extraction was carried out for Gaussian noise images with various coefficients,and then a comprehensive analysis and comparison is carried on with the other results of information extraction using the classical integral order operator,such as Roberts,Prewitt,Sobel and so on.Experiments demonstrate that the proposed approach can not only extract high-frequency edge information,but also extract texture information of smooth area and to a great extent,enjoys a certain anti-noise performance for the Gaussian noise.

关 键 词：高斯噪声 抗噪性 图像信息提取 分数阶微分 

分 类 号：TN911.73[电子电信—通信与信息系统]