基于可分解马尔科夫网的极端椒盐噪声图像滤波

Filter of Image with Polarizer Salt and Pepper Noise Based on DMN

  • 摘要: 提出了一种基于可分解马尔科夫网(decomposableMarkovnetworks,DMN)的极端椒盐噪声的均值滤波方法,指出其网络节点的阈值衰减特性和网络节点的连接特性具有很好的对椒盐噪声污染图像的噪声定位的作用,并提出一种在该网络控制下,只对与噪声相关的像素进行均值计算以替代噪声像素的亚均值滤波算法,实现了图像的较强自适应滤波。实验表明,本文方法具有良好的滤波性能。

     

    Abstract: This paper studies the filter of image with polarizer salt and pepper noise based on DMN. And it is pointed that there are abilities of fine distinction of pixel in image due to the features of decaying of threshold value, and linking of node of networks. And therefore the approaches are posed for image filtering under the control of DMN. The algorithm only calculates the mean of pixels that are related noise, and place the noise pixel. And what is called sub-mean filter algorithm. It is carried out that noise image is filtered by stubborn self-adapting method. The experimental results show that the method has better behavior of filter.

     

/

返回文章
返回