Abstract
Image enhancement of low-resolution images can be done through methods such as interpolation, super-resolution using multiple video frames, and example-based super-resolution. Example-based super-resolution, in particular, is suited to images that have a strong prior (for those frameworks that work on only a single image, it is more like image restoration than traditional, multiframe super-resolution). For example, hallucination and Markov random field (MRF) methods use examples drawn from the same domain as the image being enhanced to determine what the missing high-frequency information is likely to be. We propose to use even stronger prior information by extending MRF-based super-resolution to use adaptive observation and transition functions, that is, to make these functions region-dependent. We show with face images how we can adapt the modeling for each image patch so as to improve the resolution.
Highlights
Work on enhancing low-resolution images addressed increasing the resolution of the image without any specific outside information related to the image domain
To justify the need for having a full-Markov random field (MRF) instead of just local observation functions, we show in Table 1 the difference that having transition functions included between neighboring patches provides
We see that we can, on average, improve the resolution of the face by using MRFs whose φi, φG(i), or ψi function is adapted as indicated. This is most notable with φi adapted to its neighborhood, which reduced the mean squared error (MSE) of bilinear interpolation by 13.6%, as opposed to just 7.7% for the baseline MRF
Summary
Work on enhancing low-resolution images addressed increasing the resolution of the image without any specific outside information related to the image domain. Example-based super-resolution [4] uses the known characteristics of this domain (i.e., the prior distribution) to perform specialized enhancement They learn the priors from a database of high-resolution images from the same domain (this is in contrast to priors defined by hand [3]). Instead of using the standard method of having a single global observation function φ and a single global transition function ψ, we show how to adapt them for each patch in the face This differs from [4] in that there is a strong prior for each respective patch in the MRF.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
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 callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
