工作汇报.pptVIP

工作汇报 11月30日至12月6日 期间主要工作 按照老师要求阅读了以下几篇论文 《two-stage image denoising by principal component analysis with local pixel grouping》 《image denoising by sparse 3-D transform-domain collaborative filtering》 考虑在现有优化模型中加入新的约束项 two-stage image denoising by principal component analysis with local pixel grouping LPG-PCA denoising algorithm modeling of spatially adaptive PCA denoising 一个像素点通过2个特性进行描述：空间位置已经像素点的强度。即图像的局部结构通过不同强度的像素点及其周围的像素点的集合表示的。图像的结构信息主要通过边缘结构传达，所以在降噪中一个重要的要求就是保持图像的边缘结构不受损坏。为了这个目的，我们将每个像素点及其周围的点（nearest neighbors）看做一个向量来进行降噪，而不是对单个像素点进行处理。 two-stage image denoising by principal component analysis with local pixel grouping two-stage image denoising by principal component analysis with local pixel grouping local pixel grouping(LPG) in this paper we use the block matching method as the grouping method because of it is the simplest yet very efficient. two-stage image denoising by principal component analysis with local pixel grouping two-stage image denoising by principal component analysis with local pixel grouping LPG-PCA based denoising two-stage image denoising by principal component analysis with local pixel grouping two-stage image denoising by principal component analysis with local pixel grouping 之后对协方差矩阵进行SVD分解 two-stage image denoising by principal component analysis with local pixel grouping two-stage image denoising by principal component analysis with local pixel grouping two-stage image denoising by principal component analysis with local pixel grouping two-stage image denoising by principal component analysis with local pixel grouping image denoising by sparse 3-D transform-domain collaborative filtering 算法的主要思路，将图片用一定大小的扫描框进行扫描，将获得的图像块聚类，将每一类数据按一定的顺序排列成一个数据集（cube），之后对于每个数据集进行联合滤波，即通过一定变换将数据投影到其他平面进行滤波，之后再进行反变换，获得需要的去噪图像。 image denoising by sparse 3-D transform-domain collaborative filtering 由于图像块存在重叠的情况，所以可能需要在图像复原的过程中进行一个加权求和。 image denoising by sparse 3-D transform-domain collaborative filtering image denoising by sparse 3-D tran