[M4IF Technotes] SA-DCT

[Andrew Kinane]

Hello,
When computing the SA-DCT of an 8x8 block, each row of object pixels are
shifted horizontally and aligned to the left based on the alpha plane
information. Each row is then transformed with an N-point DCT depending
on the length of the row. This process is repeated vertically on the
resultant data. I know it doesnt matter if the horizontal or vertical
transform comes first, and I undestand the general idea no problem. But
what happens if the alpha block looks like the following ('1' = part of
object, '0' = outside object)
0    0    0    0    0    0    0    0
0    1    0    0    1    1    0    0            // the important row!
0    1    1    1    1    1    0    0
0    0    1    1    1    1    1    0
0    0    0    1    1    1    1    0
0    0    0    1    1    1    1    0
0    0    1    1    1    1    1    0
0    0    0    0    0    0    0    0
After shifting left, are we left with:
0    0    0    0    0    0    0    0
1    0    0    1    1    0    0    0            // What transform here?
1    1    1    1    1    0    0    0            // 5-point DCT
1    1    1    1    1    0    0    0            // 5-point DCT
1    1    1    1    0    0    0    0            // 4-point DCT
1    1    1    1    0    0    0    0            // 4-point DCT
1    1    1    1    1    0    0    0            // 5-point DCT
0    0    0    0    0    0    0    0
What happens with the second row? Are there two transforms performed - a
1-point (albeit trivial) and a 2-point? Are there any other
complications involved with the SA-DCT?
Regards,
Andrew
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/mp4-tech/attachments/20030204/82a4d073/attachment.html