Abstract
Two novel approaches are proposed aimed at reducing computation and memory for the fast rate allocation algorithm in JPEG2000. We propose a simplification scheme to facilitate calculation of the rate-distortion slopes based on some approximate treatments by means of logarithm expression. The other approach is fast truncation of the code stream, which can find all the truncation points for every code-block through a round search. Moreover, it need not store the rate-distortion slopes and can speed up the rate allocation. Experimental results show that it maintains high image quality and high truncation accuracy comparable to the JPEG2000 rate control.
| Original language | English |
|---|---|
| Pages (from-to) | 993-998 |
| Number of pages | 6 |
| Journal | IEEE Transactions on Consumer Electronics |
| Volume | 51 |
| Issue number | 3 |
| DOIs | |
| State | Published - Aug 2005 |
Keywords
- Coding pass
- Layer
- Rate-distortion slope
- Truncation point