Abstract
Systems and/or methods that factor large-scale repeated content within and/or among images can include devices and components that factor received or acquired images into epitomes that include all the content of the received or acquired images and transform maps that encode how to construct a facsimile or a close approximation of the image by selecting transformed regions from the epitomes. Through use of both the epitomes and the transform maps in conjunction, a facsimile or a close approximation of the input image can be reconstructed and displayed.
| Original language | English |
|---|---|
| Patent number | US8204338B2 |
| Filing date | 14/02/08 |
| Publication status | Published - 19 Jun 2012 |