Multi-frame Image Super-resolution by Total-variation Regularization
Zhigang Xu, Xiuqin Su, Zhanpeng Zhang · Institutional Repository of Xi'an Institute of Optics and Fine Mechanics, Chinese Academy of Sciences (Xian Institute of Optics and Precision Mechanics) · 2012
Multi-frame image Super-resolution (SR) reconstruction needs to deal with the recovery of a single high-resolution image from a set of low quality images. Recently, there has been a great deal of work developing multi-frame SR algorithms. In this study, we proposed an efficient SR regularization algorithm. This approach combines the ideas of bilateral filtering and the beyond-digital-total-variation model. A new regularization norm was presented, termed as locally digital bilateral-total-variation, to keep edges and more details. Experimental results were introduced to illustrate the effectiveness of the proposed algorithm. Performance analysis shows that our method is superior to similar existing methods, the peak signal-to-noise ratio improved about 0.1-1.0 dB.