Information Technology Reference
In-Depth Information
the input frame, the upscaled frame using single frame bicubic interpolation, 3-D
ISKR, and MASK at a large scale. Next, we estimated an intermediate frame at time
t
= 5
.
5 for both Spin-Calendar and Texas sequences by 3-D ISKR and MASK, and
the results are shown in Fig. 15. The intermediate frames are also spatially upscaled
(a) Input
(b) Bicubic interpolation
(c) 3-D ISKR
(d) MASK
(e) Input
(f) Bicubic interpolation
(g) 3-D ISKR
(h) MASK
Fig. 14
Spatial upscaling of Texas video sequence: (a) the input frame at
t
= 5, (b)-(d) the
upscaled video frames by bicubic interpolation, 3-D ISKR, and MASK, respectively. (e)-
(h) Enlarged images of the input frame and the upscaled frames by cubic interpolation, 3-D
ISKR, and MASK, respectively.
