Infrared Image Recovery from Visible Image by Using Multi-scale and Multi-view Sparse Representation

Methods based on sparse coding have been successfully used in IR image Super-resolution (SR). However, existing sparse coding based SR usually encounters several problems. To overcome these problems, we propose a novel IR image SR method. First, we combine the information from multi-sensors to improve the resolution of the IR image. Secondly, we use multi-scale patches to represent the IR image more accurately. Finally, we use different kinds of feature vectors to represent the IR image. Extensive experiments validate that using the proposed method yields better results in terms of quantitation and visual perception than many state-of-the-art algorithms.