摘要： 提出一种基于多帧低分辨图像融合的超分辨率图像恢复算法。将多帧低分辨图像融合成一帧与高分辨图像分辨率一致的图像，并对其中一幅低分辨图像插值，形成迭代恢复算法的初始值，在此基础上以Tikhonov正则化方法求解原始高分辨图像。分析和实验结果表明，该算法具有良好的鲁棒性，并且计算速度较快。

关键词: 图像恢复, 超分辨率, 图像融合

Abstract: This paper offers a Super-Resolution(SR) image restoration algorithm based on Low-Resolution(LR) image fusion. The algorithm fuses a set of LR images to form an image with same resolution as the High-Resolution(HR) image, and interposes a LR image as the initialization guess for the iterative restoration algorithm. Both the above images are used to find the original HR image by the Tikhonov regularization algorithm. Analysis and simulation results confirm that the algorithm has good robustness and faster computation.

Key words: image restoration, Super-Resolution(SR), image fusion

中图分类号: 

• TP391.4