SH7785 Datasheet, PDF (1050/1692 Pages) Renesas Technology Corp – 32-Bit RISC Microcomputer

20. Graphics Data Translation Accelerator (GDTA)
(1) Estimated Image Generation Function
The following shows an outline of the estimated image generation function.
Output frame
Y pointer (current)
Output frame
U pointer (current)
Output frame
V pointer (current)
Past frame
Y pointer
Past frame
U pointer
Past frame
V pointer
Future frame
Y pointer
Future frame
U pointer
Future frame
V pointer
DDR2-SDRAM
(1) Calculation of output position (first row) (2) Calculation of output position (nth row)
Base point
Output
target position
Y (current)
(mbcol, mbrow)
U (current)
V (current)
Image data for one frame
mbcol
(9) Estimated
image writing
Base point
Frame width
Y padding
Target position (mbcol, mbrow)
Comparison point
Y (past)
(+Recon_right, down)
U (past)
V (past)
(5) Data
reading
Y (future)
U (future)
V (future)
Recon_right (negative value indicates reverse direction)
Output
(current)
Y: 16 × 16 dot
U: 8 × 8 dot
V: 8 × 8 dot
Input
Hardware performs processing
image
as is even when in padding part
(half-pixel)
Y: 17 × 17 dot
For half-pixel processing,
get data for a block with one
U: 9 × 9 dot _{extra dot according to motion}
V: 9 × 9 dot vector values
Each frame pointer should point to an address on a 16-byte boundary
for Y and 8-byte boundary for U and V.
Correction
case 1
Correction
case 2
Correction
case 3
Input data for one macroblock
(3) Calculation of input position (first row)
(4) Calculation of input position (nth row)
Buffer RAM 1
AB
× 17
(6) Half-pixel correction processing
ED
Correction case 4
IDCT data
Y0 (128 bytes)
Y1 (128 bytes)
Y2 (128 bytes)
× 17
Y: 17 × 17 dot
U: 9 × 9 dot
V: 9 × 9 dot
(7) IDCT data
reading
Y3 (128 bytes)
U (128 bytes)
V (128 bytes)
A' B'
Correction processing result (Y: 16 × 16, U/V: 8 × 8)
× 16
ab
× 16
(8) Estimated-image generation
A'' B'
× 16
C' D'
cd
C' D'
× 16
Y: 16 × 16 dot
U: 8 × 8 dot
V: 8 × 8 dot
Data after half-pixel
correction
× 16
Y: 16 × 16 dot
U: 8 × 8 dot
V: 8 × 8 dot
IDCT data
8-bit saturation
calculation
× 16
(0 ≤ x ≤ 255)
Y: 16 × 16 dot
U: 8 × 8 dot
V: 8 × 8 dot
Data to be written back to the current target position
(result of forward macroblock processing)
Figure 20.6 Outline of Estimated Image Generation Function
Rev.1.00 Jan. 10, 2008 Page 1020 of 1658
REJ09B0261-0100