一种新的基于SVD图像降噪方法的探究

奇异值分解(SVD)在图像处理中具有极其重要的应用.针对传统SVD降噪适应性差的缺点,提出了在得知噪音方差的前提下,以基于SVD和能量最小原则图像自适应降噪为参考,在保持相近峰值信噪比的同时,将图像降噪的适应范围进一步推广.并且运行时间得到了有效缩减,为处理大型图像问题提供了可能性.数值试验结果表明,同种情况下分块降噪效果更佳.

A New Research For The Method Of Based On SVD Image Denoising

Singular value decomposition (SVD) has very important applications in image processing.Aiming at the SVD of the denoising unadaptable faults, this paper puts forward the noise variance based on the premise, SVD and minimum energy principle of adaptive image denoising for reference , in keep similar peak SNR at the same time, the scope of the image denoising of the methods to further promote. And running time is also effectively reduced, provides the possibility for the treatment of large image problem. Numerical results show that under the same conditions, block method is better.