Page 131 - Video Coding for Mobile Communications Efficiency, Complexity, and Resilience
P. 131
108 Chapter 4. Basic Motion Estimation Techniques
Foreman @ 8.33 f.p.s.
33
SSD
SAD
32
NCCF
31
30
PSNR Y (dB) 29
28
27
26
25
1 2 3 4 5 6 7 8 9 10
Frame
Figure 4.3: Reconstruction quality of SSD, SAD, and NCCF
Table 4.1: Computational complexity of SSD, SAD, and NCCF for an
N × N block
SAD SSD NCCF
|·| N 2 – –
− N 2 N 2 -
2
2
2
+ N − 1 N − 1 3(N − 1)
2
× – N 2 3N +1
÷ – – 1
√ – – 2
4.6.2 Block Size
Another important parameter of the BMA is the block size. Figure 4.4 shows
the performance of the BMA with two di erent sizes, 8 × 8 and 16 × 16. It can
be seen in Figure 4.4(a) that a smaller block size achieves better prediction
quality. This is due to a number of reasons. A smaller block size reduces
the e ect of the accuracy problem. In other words, with a smaller block size,
there is less possibility that the block will contain di erent objects moving in
