generic-library/vpx
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Coding con-zero count rather than EOB for coeffs

Browse Source 
This patch revamps the entropy coding of coefficients to code first
a non-zero count per coded block and correspondingly remove the EOB
token from the token set.

STATUS:
Main encode/decode code achieving encode/decode sync - done.
Forward and backward probability updates to the nzcs - done.
Rd costing updates for nzcs - done.
Note: The dynamic progrmaming apporach used in trellis quantization
is not exactly compatible with nzcs. A suboptimal approach has been
used instead where branch costs are updated to account for changes
in the nzcs.

TODO:
Training the default probs/counts for nzcs

Change-Id: I951bc1e22f47885077a7453a09b0493daa77883d
This commit is contained in:
Deb Mukherjee
2013-02-20 10:16:24 -08:00
parent a9961fa819
commit eb6ef2417f
25 changed files with 3330 additions and 384 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
vp9/encoder/vp9_segmentation.c
View File

@@ -219,10 +219,8 @@ static void count_segs(VP9_COMP *cpi,
const int segment_id = mi->mbmi.segment_id;
xd->mode_info_context = mi;
xd->mb_to_top_edge = -((mb_row * 16) << 3);
xd->mb_to_left_edge = -((mb_col * 16) << 3);
xd->mb_to_bottom_edge = ((cm->mb_rows - mb_size - mb_row) * 16) << 3;
xd->mb_to_right_edge = ((cm->mb_cols - mb_size - mb_col) * 16) << 3;
set_mb_row(cm, xd, mb_row, mb_size);
set_mb_col(cm, xd, mb_col, mb_size);
// Count the number of hits on each segment with no prediction
no_pred_segcounts[segment_id]++;