Abstract

Super resolution aims to produce a high-resolution image from a set of low-resolution images. Varieties of techniques have been proposed for this task. Among them, optimization methods have been used frequently to combine information in low-resolution images. Optimization methods minimize an error function usually consisting two terms. The first one models similarity between high and low resolution images while the second term intends to reduce the amount of noise in the obtained image. Optimization methods typically minimize error function globally. To reduce the blur effect on the high-resolution image, coefficient of the second term should be low enough. On the other hand, for noise reduction, high coefficient is desirable. In this paper we proposed a new method to locally minimize the error function based on region of interest (ROI) selection. With this approach, the first term is reduced in the image's regions with high amount of information while the second term is minimized only in the smooth areas.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call