一种超分辨率图像重构算法研究

Algorithm on Super-resolution Image Reconstruction

为了从低分辨率序列图像中重构出高分辨率的图像,该文在Robust超分辨率图像重构算法中引入了正则化因子,提出了一种新的超分辨率图像重构算法,它不仅消除了低分辨率图像中的奇变信息(bias),而且增强了抑制超分辨率图像重构噪声的能力。实验结果表明,该文提出的算法具有更好的效果(MSE值更小)。

In order to extrapolate a super-resolution image from a low-resolution image sequence,this paper proposes a new algorithm to reconstruct super -resolution image by using a regularization factor into the traditional Robust algorithm.This new algorithm not only embodies the robust effect to remove the bias but also enhances the ability to remove noise in the process of super-resolution image reconstruction.Experimental results show that the new algorithm brought forward in this paper has better performance and less MSE than traditional algorithm.