摘要
灰度图像消噪方法一般是借助于图像的平滑处理技术如均值、中值消噪、数学形态学消噪。这些常规消噪方法都是在空间域对像素的邻域进行某种平滑或积分处理,反映到频域就是在图象的高频区域进行分量抑制或衰减,从而不可避免地导致图像轮廓或边缘区域变得模糊。基于比特平面分解的图像消噪方法,是依据分解所得的各比特平面对原图像边缘信息的贡献,并不对高位的比特平面进行平滑消噪,使得消噪后图像的轮廓或边缘的清晰度较常规平滑消噪有所改善。
The denoising methods of a gray-scale image are commonly used by smoothing out the image such as mean value, middle value and closing operation in morphology. These denoising methods always average or integrate the near domains of all pixels in a grayscale image. That is to say these methods restrain or attenuate the values of coefficients in an image＇s frequency domain in result ofbluring the image, and especially the edge of the image. A denoising method of images based on their decomposed bit planes has less degree of bluring an image, and especially the edge of the image than others without denoising the main bit planes which contain mostly the information of the edge of the image.
出处
《计算机工程与设计》
CSCD
北大核心
2007年第2期392-394,共3页
Computer Engineering and Design
关键词
比特平面
图像分解
图像消噪
边缘提取
平滑
模糊
bit plane
image decomposing
image denoising
edge extraction
averaging
bluring
