光学学报, 2014, 34 (s2): s211004, 网络出版: 2014-12-02  

基于阿尔法均值滤波和信息散度的图像滤波研究

Filter of the Image Based on Alpha-Trimmed Mean Filter and Information Divergence
作者单位
浙江工业大学信息工程学院, 浙江 杭州 310014
摘要
光学图像感染噪声是一种非常常见的现象,为了克服传统图像滤波方法不可避免带来的图像边缘模糊和奇异值影响,提出了一种基于阿尔法均值滤波算法和信息散度的图像自适应滤波算法。该算法充分利用了信息散度的特性和阿尔法均值滤波算法的优点,采用信息散度自适应的确定滤波窗口的加权系数值。实验结果表明,该算法可动态的确定滤波窗口内像素点的取值,具有较好的自适应性,同时与现有的中值滤波和均值滤波算法相比,对于受到高斯噪声、椒盐噪声以及混合噪声感染的图像,具有比较好的滤波效果,并可以很好地保护图像细节信息,具有提高图像清晰度的功能。
Abstract
The optical image disturbed by noise is a very common phenomenon. In order to overcome the fuzzy effects among the imaging edges and those of singular values induced by the traditional image filtering methods, an adaptive filtering algorithm is proposed on the basis of Alpha-trimmed averaging filter and information divergence theorem. The algorithm makes full use of the advantages of the Alpha-trimmed averaging filter and information divergence. It can adaptively adjust the filter window′s coefficients according to the information divergence. The experimental results show that the high adaptive characteristic due to the dynamically determination for the pixel value in the filtering window. Meanwhile, for the image corrupted by Gaussian noise, salt and pepper noise or mixed noise, the proposed algorithm not only accounts for much more advantages in filtering capacities, but also enhances the detailed distinguishing features of images and the corresponding resolutions while comparing with the present middle and average ranged filtering algorithms, respectively.

常丽萍, 李伽, 施朝霞, 郭淑琴. 基于阿尔法均值滤波和信息散度的图像滤波研究[J]. 光学学报, 2014, 34(s2): s211004. Chang Liping, Li Jia, Shi Zhaoxia, Guo Shuqin. Filter of the Image Based on Alpha-Trimmed Mean Filter and Information Divergence[J]. Acta Optica Sinica, 2014, 34(s2): s211004.

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!