期刊名称：International Journal of Computer Science and Network Security
印刷版ISSN：1738-7906
出版年度：2017
卷号：17
期号：10
页码：256-261
出版社：International Journal of Computer Science and Network Security
摘要：Infrared (IR) images are significant in several major fields such as security, military and landform determination. However, due to physical limitations of the precision optics and expensive image sensors, these images tend to have poor resolution. This paper presents a Single-image Super-Resolution (SISR) algorithm for IR thermal images which effectively reconstructs High-resolution (HR) image from its low-resolution (LR) counterpart without an external database. In this method, the training data set is built from in-place self-examples generated by a bi-pyramid of recursively scaled and subsequently interpolated image patches. The relation between self-example patch-pairs is learned by a regression operator represented as a matrix and used as a prior to super-resolve LR infrared thermal images. Subjective and objective evaluation of the proposed algorithm validates the efficiency of the proposed algorithm.
