| TEMPORAL SCALABILITY MODEL |
|
|
|
|
| • | Each focal object is represented by a
focal window described by |
||
| three parameters: |
|||
|
|
|
|
|
|
| By an associated frame (Fn), region |
|||||
| geometry (Ân) (such as if it
is an |
|||||
| rectangular region, then shape-type, x and |
|||||
| y coordinates of a corner, and height and |
|||||
| width), and focal acuity parameters (An) |
|||||
| (such as relative level of spatial bit |
|||||
| allocation or desired temporal resolution). |
|||||
| = |
| Â |
| O |
| { |
| F |
| , |
| , |
| A |
| } |
| n |
| n |
| n |
|
|
| • | Frame Synthesis |
|
|
|
|
|
| The superposition operation is non- |
||||
| commutative.
This is performed by sorting |
||||
| the focal regions in order of their skip-rate |
||||
| (the focal region with maximum skip rate is |
||||
| committed
first). |
||||
| é |
| ù |
| = |
| Â |
| F |
| F |
| t |
| t |
| t |
| r |
| û |
| sup |
