基于稀疏表示的图像超分辨率重建

Image Super-Resolution Reconstruction Based on Sparse Representation

针对单幅低分辨率灰度图像,提出一种基于稀疏表示和多成分字典学习的超分辨率重建算法。对于输入的低分辨率图像,通过训练两个字典,并根据每一个图像块在高分辨率字典上的稀疏表示产生高分辨率图像输出。为了提高重建图像的质量,对稀疏字典设计方法进行改进,采用K-SVD方法对图像进行MCA分层,提取出图像中的Texture和Cartoon部分,用于字典的学习和超分辨率的重建。仿真实验结果表明,改进的算法在图像信噪比上有所提高。

Proposes a super-resolution reconstruction approach of single gray image based on sparse representation and multi component-dictionary learning. For each patch of the low-resolution input image, considers a sparse representation to train two dictionaries and then uses the coefficients of this representation to generate the high-resolution output image. In order to improve the quality of the reconstructed image, improves the design method of the sparse dictionary using the method of K-SVD layered image in MCA to extract the components of Tex-ture and Cartoon in the image for dictionary learning and super-resolution reconstruction. The results of simulation experiment show that the method leads to an improvement in PSNR.