基于双正则化参数的在线字典学习超分辨率重建

基于学习的单图超分辨率重建算法能获得较好的超分效果，但存在重建图像伪影较为明显的问题。为解决这一问题，提出了一种基于双正则化参数的在线字典学习超分辨率重建算法。在字典学习过程中运用在线字典学习方法（Online Dictionary Learning，ODL），并在稀疏字典生成阶段和图像重建阶段分别设置了两个不同的正则化参数。实验中生成的目标高分辨率图像PSNR比经典的稀疏编码超分方法（Sparse Coding Super-Resolution，SCSR）平均提高了0.39dB，较好地恢复图像边缘锐度和纹理细节的同时有效地抑制了伪影。ODL和双正则化参数的引入，提高了字典训练的精度，使字典训练和图像重建阶段的稀疏系数独立可调，实验中能够有效地消除伪影，提升了超分辨率重建的效果。

Image Super-Resolution Based on Online Dictionary Learning with Two Regularization Parameters

The performance of some learning-based super-resolution methods is promising, but some obvious artifacts appear in the reconstruction images. In order to solve this problem, this paper presents a novel super-resolution algorithm based on Online Dictionary Learning (ODL) with two regularization parameters. It employs ODL in the dictionary learning procedure. Then the algorithm sets two regularization parameters in the procedures of dictionary learning and image reconstruction. In the experiments, the PSNRs of the new method are 0.39 dB higher than the state-of-the-art Sparse Coding Super-Resolution (SCSR) in average. It can eliminate the artifacts while recovering the edge sharpness and the texture details efficiently. With the introduction of ODL and two regularization parameters, it promotes the dictionary training accuracy and makes the sparse coefficients in dictionary learning and image reconstruction adjustable separately. The experiments show that the artifacts are eliminated effectively. It promotes the final effect of super-resolution reconstruction well.