摘要
邻域平均法(NAM)应用于图像去噪时能够得到较好的平滑效果,但图像的细节信息丢失较多;基于小波变换的阈值图像去噪方法能够较好地保持图像的细节信息,但是平滑效果不理想.对这两种去噪方法所得到的图像进行小波变换,然后在小波域再按照一定的融合规则进行融合处理,得到去噪效果较好的图像.实验结果表明,融合后的图像能够较好地去除图像的噪声,具有较好的视觉效果和较高的峰值信噪比.
Image denoising based on Neighborhood Averaging Method(NAM) imparts better smooth effection,but the detail of the image is lost greatly. Image denoising based on wavelet transform can keep the detail of image better, but the effect of image smoothing is worse. When the two results of these denoising methods are fused according to certain fusion rules based on wavelet transform we can get a new better denoising image. And the experimental result shows that the new denoising image can keep the image detail better,reduce the image denoising effectively and get a high peak signal-to-noise ratio(PSNR).
出处
《泉州师范学院学报》
2007年第2期49-51,61,共4页
Journal of Quanzhou Normal University
关键词
图像去噪
图像融合
小波变换
image denoising
image fusion
wavelet transform