基于图像旋转和分块的奇异值分解图像去噪

奇异值分解在图像处理中具有重要应用，针对奇异值分解在图像去噪中存在的问题，提出了分块旋转奇异值分解图像去噪新方法。首先分析了奇异值分解在图像去噪中的应用及对非竖直水平方向信息的图像在去噪时存在的问题；采用图像旋转的方法，将非竖直水平方向信息的图像变为竖直或水平方向信息的图像，较好克服了奇异值分解在图像去噪中的内在问题；对于图像具有的多方向性信息，通过图像分块获得各个图像块的方向性信息，分别对每个图像块进行旋转奇异值分解去噪，获得了最佳的奇异值分解图像去噪效果。仿真结果表明，与传统的奇异值分解去噪算法相比，

Image denoising based on SVD using image rotation and block

Singular value decomposition (SVD)was important in image processing.Due to the problem of SVD in image denoising,a method of block-based and rotation SVD denoising was presented.Firstly,the application in image denoising using SVD and the problem in non vertical and horizontal information image denoising using SVD were analyzed.And the image of non vertical and horizontal information was rotated to the image of vertical or horizontal information using image rotation.Then the inherent problem was overcome.The best result of SVD denoising was obtained though block-based direction information when the image has lots of direction information and the image block denoising using rotation based on SVD.Experiments show that the proposed method of block-based and rotation SVD denoising was effective compared to the traditional SVD denoising,the visual quality and PSNR were improved substantially.