Use 8x8 DCT transform for I8X8 prediction mode
authorJingning Han <jingning@google.com>
Wed, 1 Aug 2012 17:18:25 +0000 (10:18 -0700)
committerJingning Han <jingning@google.com>
Thu, 2 Aug 2012 16:09:17 +0000 (09:09 -0700)
commitc7846ebc3438aa3e1611398567bf950d0656c590
treedbec3e20a943343731a1ba72aaa5fc376a97d85d
parent171b4ae1575ccfcf02dcaebd49a099c3a23fa0eb
Use 8x8 DCT transform for I8X8 prediction mode

Apply 2D-DCT transform of dimension 8x8 to encode prediction
residuals of I8X8 mode.
Brought back block type 3 probability context model for 8x8 tokens,
which is used for the coefficients of Y blocks in I8x8 modes. The
coefficient costs estimate of I8X8 mode in rate-distortion is also
changed appropriately.
Performance results:
derf:   0.246
yt:     0.114
std-hd: 0.730
hd:     0.670

Change-Id: If1d970eeb4e1827c9f0d2c5b27d33089b347ea27
configure
vp8/common/default_coef_probs.h
vp8/common/entropy.h
vp8/decoder/decodframe.c
vp8/decoder/detokenize.c
vp8/encoder/encodeintra.c
vp8/encoder/rdopt.c
vp8/encoder/tokenize.c