基于图像去噪的最大均匀平滑法的改进

Improvement of Maximum Uniform Smoothing Method Based on Image Denoising

为了改善图像的质量,在去除图像噪声点的同时,保护人们所感兴趣的区域,提出了一种改进算法。利用邻域平均法的思想,基于对像素周围点求和平均后替换中心像素值的方式,提出了一组新的掩膜措施。根据模板像素与中心点的不同距离,设定新的加权模板,利用matlab对图像各个像素点进行分类判断,之后针对图像中的背景、边线特征的不同,选择相对应的模板,对图像进行滤波处理。对该算法与传统的均值滤波、中值滤波使用图像平滑的客观评价方法进行定量指标比较。实验结果表明,对于有不同结构或内容的图像,该算法在图像平滑后的效果更好,尤其是在图像比较复杂或细节比较多的情况下,图像能够在平滑后显示更多的边缘或纹理。

In order to improve image quality,an improved algorithm is proposed to protect the areas of interest while removing the noise points.According to the idea of neighborhood average method,based on the method of replacing the center pixel value with the average sum of the points around the pixel,a new set of mask measures is proposed.A new weighted template is set according to the different distance between the template pixel and the center point,and each pixel point of the image is classified and judged by matlab.After that,the corresponding template is selected to filter the image according to the different background and edge features in the image.The proposed algorithm is compared with the traditional mean filter and median filter in quantitative index by the objective evaluation method of image smoothness.Experiment shows that for images with different structures or contents,the proposed algorithm is more excellent after image smoothing,especially in the case of more complex image or more details,the image can show more edges or textures after smoothing.