基于动态窗口的自适应中值滤波算法  

DYNAMIC-WINDOW-BASED ADAPTIVE MEDIAN FILTERING ALGORITHM

作　　者：周华[1] 

机构地区：[1]德州学院计算机系,山东德州253023

出　　处：《计算机应用与软件》2011年第7期141-143,166,共4页Computer Applications and Software

基　　金：山东省优秀中青年科学家科研奖励基金(BS2009SF014)

摘　　要：针对中值滤波算法对于高密度噪声图像以及纹理细腻图像的边缘处理能力欠佳的缺陷,提出一种基于动态窗口的自适应中值滤波算法。新算法根据噪声点与周围信息的关联程度将噪声点滤波值进行调整,从而更好地处理图像的细节部份。新算法中的自适应策略加强了滤波算法的去噪性能,使其对于含有任意噪声密度的图像也能很好地进行噪声滤除。通过仿真分析,新算法对于细节丰富的图像以及高密度噪声的图像滤波效果良好,有效地提高图像的峰值信噪比,其去噪效果相比其他方法更加优秀。Aiming at the drawbacks of the poor edge handling capacity for high-density noised images and fine-texture images of the median filtering algorithm,the thesis has proposed an adaptive median filtering algorithm based on the dynamic window.The new algorithm adapts the filtration value at the point of the noise in accordance with how closely related the point of noise to its surrounding information in order to better deal with the minor parts of images.The adaptive strategy of the new algorithm upgrades the de-noising performance of the filtering algorithm so that it can well filter noises from an arbitrary noise density image.Simulation shows the novel algorithm works well on intricate images and high-density noised images,effectively improves the peak signal-noise ratio of images,and is better at de-noising than other algorithms.

关 键 词：椒盐噪声 噪声检测 动态窗口 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]