【摘要】 为了有效的去除图像中的脉冲噪声,提出了一种新的噪声检测与噪声去除的方法,它在充分分析了噪声的分布特性的基础上,对图像中的噪声及图像的细节进行多方向检测,进而有效的分辨出噪声像素点与非噪声像素点。同时在传统的中值滤波的基础上结合邻域均值加权的思想对检测到的图像噪声点进行处理。在Matlab上的实验结果表明,该方法能有效地检测出噪声点,同时也在一定程度上较好的保存了图像的边缘信息。更多还原

【Abstract】 In order to effectively remove the impulse noise of the images,a new approach about noise detection and noise reduction is presented,it is in the full analysis of distinguishing noise pixels and non-pixel noise by multi-direction detection,the effective resolution to the noise removing comes from the traditional median filter,it is on the basis of the weighted average neighborhood to deal with the noise. It can conclude that the method can effectively detect the noise,and also to some extent better preserve the image of the edge information by the trial in Matlab.更多还原