各向异性扩散-中值滤波在红外图像处理中的应用  

Anisotropic Diffusion-Median Filter for Infrared Image

作　　者：刘潮东[1] 史忠科[1] 

机构地区：[1]西北工业大学自动化学院,西安710072

出　　处：《弹箭与制导学报》2006年第3期198-200,共3页Journal of Projectiles,Rockets,Missiles and Guidance

摘　　要：在低信噪比图像处理中,为了在图像去噪时更好地保留边缘,提出了一种各向异性扩散-中值滤波方法。这种方法包含两个步骤:首先,为了使滤波器在抑制噪声时具有鲁棒的边缘保持能力,该方法采用各向异性扩散滤波对图像进行处理,其中,各向异性扩散滤波中的扩散函数选用Tukey函数;然后,为了去除各向异性扩散滤波后图像中的脉冲噪声,采用中值滤波对图像做进一步的处理,从而使这种滤波方法在抑制噪声的同时使边缘模糊达到最小。从仿真结果可以看出,文中的方法具有良好的噪声抑制和边缘保持能力。In order to preserve edge while removing noise, an anisotropic diffusion median filter is proposed for low-signal-to-noise image. The method includes two steps： firstly, to achieve much better noise suppression with robust capability of edge-preserving, anisotropic diffusion is employed to image processing, whose diffusivity function is Tukey function; secondly, in order to remove impulse noise, median filter is employed for the processed image. So this algorithm can achieve much better noise suppression with minimum edge blurring. The experimental result shows that our has good capability in noise suppression and edge-preserving.

关 键 词：各向异性扩散 中值滤波 扩散函数 偏微分方程 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]