Cryptography Reference
In-Depth Information
38.0
37.5
37.0
36.5
36.0
35.5
35.0
34.5
34.0
33.5
33.0
32.5
32.0
t
+2D
J
SVM3.0
STool
31.5
31.0
30.5
30.0
Spatio-Temporal and Quality Resolution
Fig. 3.19.
Comparison of wavelet schemes with AVC-based SVC for the Harbour
sequence [32].
3.3.6 Performance Comparison
The scalable extension of AVC/H.264 has been chosen to be the standard,
the wavelet-based approach also has many promising features and open ques-
tions. In recent reports, the VidWave software can achieve comparable per-
formance to the JSVM [24, 25, 30, 31, 32]. Figs. 3.19 and 3.20 show the
performance comparison between the AVC-based scheme (denoted as JSVM)
and the wavelet-based approach. They are extracted from the files in [32].
Fig. 3.19 is a case that favors the wavelet-based schemes. So far, the wavelet
schemes show better results on high-resolution large motion pictures. The no-
tion 4CIF is a picture format four times of the CIF size. This is about the
size of the regular TV picture. Fig. 3.20 has nearly the worst performance of
all wavelet schemes in [32].
Based on the comments and discussions in the VidWav group documents
[24, 25, 30], the wavelet-based approach has its specific strengths but it also
has weakness. Here is a brief summary. The advantages of the wavelet ap-
proach are as follows. The multi-resolution wavelets along both temporal and
spatial axes provide a natural representation of image sequences. The wavelet
video scheme provides a very large range of scalability. With the aid of em-
bedded quantization and arithmetic coding, it offers an elegant, fine-grain
SNR scalability. A typical wavelet video scheme has an open-loop structure,
and it facilitates very flexible bit stream truncations. Another advantage of
