基于纹理分析的改进型Nagao滤波器

The improved Nagao filter based on texture analysis

介绍了Nagao滤波器的基本原理,针对传统Nagao滤波器滤波后图像存在伪像以及用计算机实现处理速度慢等缺点,提出了一种基于纹理分析的改进型Nagao滤波器。通过对图像进行纹理分析,得到图像纹理复杂度指标,并用该指标作为Nagao滤波器选择模板的参数。实验分析表明,该方法可以很好地克服传统Nagao滤波器的不足,在进行图像自适应滤波时取得了比较好的效果,非常适合于对计算机处理速度要求比较高的应用。

The paper introduced the basic principle of Nagao filter, then indicates that the classical Nagao filter has some shortcomings, such as, has artifact after filtering, slow processing speed with computer and so on. So a kind of improved Nagao filter based on texture analysis is presented in this paper. Firstly, get the index of texture complexity throughout texture analysis, then using this index to select the board of Nagao filter. The experiment analysis shows that this method can overcome the shortcomings of traditional Nagao filter, and get good results during the adaptive image filtering, very suitable for the application of high computer processing speed requirement.