Digital Signal Processing Reference
In-Depth Information
60
60
50
50
40
40
30
30
20
20
10
10
10
20
30
40
50
60
10
20
30
40
50
60
(a)
(b)
60
60
30
50
50
25
40
40
20
30
30
15
20
20
10
10
10
5
5
10
15
20
25
30
10
20
30
40
50
60
10
20
30
40
50
60
(c) (d) (e)
FIGURE 6.5
:
Modulus of the SAR images of the Backhoe data obtained from a 32
32
data matrix with missing samples. (a) 2-D complete-data WFFT, (b) 2-D complete-data
APES with a 2-D filter of size 16
×
16, (c) 2-D data missing pattern, the black stripes
indicate missing samples, (d) 2-D WFFT, and (e) 2-D MAPES-CM with a 2-D filter of
size 16
×
×
16.
28 and in columns 8, 9, 10, 20, 21, 22, 26, 27 are missing (possibly due to both
angular diversity and strong interferences), resulting in 40% missing data. Fig. 6.4(e)
shows the WFFT image, which has low resolution, high sidelobes, and smeared
features. By using an initial filter matrix of size 6
6, the GAPES image is shown
in Fig. 6.4(f ), where strong artifacts around the dihedrals are readily observed.
The image reconstructed via MAPES-CM is shown in Fig. 6.4(g), which is quite
similar to the complete-data image in Fig. 6.4(c).
By observing that the missing-data algorithms developed previously estimate
the missing data samples, we can achieve better spectral estimation performance,
×
Search WWH ::
Custom Search