Information Technology Reference
In-Depth Information
by the same factor (1 : 2). Again, both 3-D ISKR and MASK produce high quality
frames in which camera noise and blocking artifacts are almost invisible while the
important contents are preserved.
(a) 3-D ISKR
(b) MASK
(c) 3-D ISKR
(d) MASK
(e) 3-D ISKR
(f) MASK
(g) 3-D ISKR
(h) MASK
Fig. 15
Spatiotemporal upscaling of Spin-Calender and Texas video sequences: (a),(c) the
estimated intermediate frames at time
t
= 5
.
5 by 3-D ISKR, (b),(d) the estimated intermediate
frames by MASK. The frames are also spatially upscaled with the upscaling factor of 1 : 2.
The images in (e)(f) and (g)(h) are the enlarged images of the upscaled frames by 3-D ISKR
and MASK, respectively.
