Adaptive super resolution for video enhancement
US8594464B2 · kind B2 · utility
Assignee
Inventor
Key dates
| Filing date | May 26, 2011 |
| Grant date | Nov 26, 2013 |
| Priority date | — |
| Expiry date | Feb 14, 2032 |
Classification
- Technology area (CPC H)Electricity
- CPC primaryH04N7/0137
- WIPO fieldAudio-visual technology
- WIPO sectorElectrical engineering
Abstract
A method of operating a computer system to reconstruct a high-resolution frame from a set of low-resolution frames of a video sequence is described. The high-resolution frame may be displayed on a high-resolution display device such as a high-definition television (HDTV). Initial estimates and priors for each of a plurality of quantities including blur kernel, motion between frames, noise level, and the estimate of the high-resolution frame are determined. The initial estimates are refined using an iterative process that iterates between estimating each of the plurality of quantities based on the current estimates for the plurality of quantities. When it is determined that a quality metric is above a threshold, further iteration is halted and the high-resolution frame is output for storage and/or display. The process may be repeated for outputting a high-resolution video sequence based on an input low-resolution video sequence.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.