Image super-resolution reconstruction algorithm based on Bayesian theory
Wenbo Zheng, Fei Deng, Shaocong Mo, Xin Jin, Yili Qu, Jiangwei Zhou, Rui Zou, Jia Shuai, Zefeng Xie, Sijie Long, Chengfeng Zheng · 2018
The Bayesian theory provides a new solution to image super-resolution reconstruction. In view of the poor robustness to noise and motion estimation in the vast majority of superresolution reconstruction algorithms. In this paper, we propose an image super-resolution reconstruction algorithm based on Bayesian representation. In the proposed algorithm, uncharted super-resolution images, motion parameters and unknown model parameters are utilized for modeling in a hierarchical Bayesian framework. We adopt degenerate distribution to derive the estimation of analytic solutions and applied the solutions to the super-resolution reconstruction which also enables the proposed algorithm robust to noises. The experimental results show that the proposed image super-resolution reconstruction algorithm based on Bayesian representation can achieve higher (or similar) performance than the state of-the-art methods.