61 
) + Ar (26) 
m that can be 
age is solved 
on 
(27) 
p size which 
(28) 
le objective 
$/12/2003» 
'2004 and 
08/01/2004, (g) cubic interpolated image, and (h) SR 
reconstructed image. 
The proposed algorithm was tested using real MODIS band-4 
images which were captured on 28, 30 December 2003 and 1, 4, 
6, 8 January 2004 respectively. We obtained these images from 
the Satellite Remote Sensing Receiving Station of Wuhan 
University. We chose two series of 50x50 regions to test the 
algorithm independently. The multi-temporal block images of 
the two regions are respectively shown in Figures 2(a)-(f) and 
Figures 3(a)-(f). 
(e) (0 
Figure 3. (a)-(f) Images respectively captured on 28/12/2003, 
30/12/2003, 01/01/2004, 04/01/2004, 06/01/2004 and 
08/01/2004, (g) cubic interpolated image, and (h) SR 
reconstructed image. 
In the experiments, the corresponding parameters were set 
as. fi ~ \ , X = 0.008 and 7V = 10. A 3x3 Gaussian blur 
kernel with unit variance was assumed and was commonly 
employed. We assume the down-sampling factors in both the 
horizontal and vertical directions have a value of 2. The cubic 
interpolated images which are shown in Figure 2(g) and Figure 
3(g) were regarded as the initial guesses of the SR images. The 
corresponding SR reconstructed results are respectively shown 
in Figure 3(h) and Figure 4(h). By visual comparison, it is seen 
that the results of the proposed SR algorithm are much clearer 
than those of the cubic interpolation algorithms, the reason for 
which is that these results fused the complementary information 
in different observed images. 
6. CONCLUSIONS 
In this paper, we have proposed a SR image reconstruction 
algorithm to multi-temporal remote sensing images. Employing 
the joint MAP framework, the proposed algorithm can 
simultaneously estimate the image registration parameters and 
the HR image. We tested this algorithm using multi-temporal 
MODIS images. Experiment results validated the proposed 
algorithm performs better than single cubic interpolation 
method. 
REFERENCES 
Conel, J. E. 1985. Calibration of data using ground-based 
spectral reflectance measurements, the Airborne Imaging 
Spectrometer Data Analysis Workshop, California, JPL 
Publication 85-41, Jet Propulsion Laboratory, Pasadena. 
Elad, M. and A. Feuer 1997. "Restoration of a single 
superresolution image from several blurred, noisy, and 
undersampled measured images." IEEE Transactions on Image 
Processing 6(12): 1646-1658. 
Elad, M. and A. Feuer 1999. "Superresolution restoration of an 
image sequence: Adaptive filtering approach." IEEE 
Transactions on Image Processing 8(3): 387-395. 
Hardie, R. C., K. J. Barnard, et al. 1997. "Joint MAP 
registration and high-resolution image estimation using a 
sequence of undersampled images." IEEE Transactions on 
Image Processing 6(12): 1621-1633. 
Hardie, R. C., T. R. Tuinstra, et al. 1997. High resolution image 
reconstruction from digital video with global and non-global 
scene motion. Proceedings of the IEEE International 
Conference on Image Processing, Santa Barbara, CA, USA, 
IEEE Comp Soc, Los Alamitos, CA, USA. 
Irani, M. and S. Peleg 1991. "Improving resolution by image 
registration." CVGIP: Graphical Models and Image Processing 
53(3): 231-239. 
Park, S. C., M. K. Park, et al. 2003. "Super-resolution image 
reconstruction: A technical overview." IEEE Signal Processing 
Magazine 20(3): 21-36. 
Patti, A. J., M. I. Sezan, et al. 1994. High-resolution image 
reconstruction from a low-resolution image sequence in the 
presence of time-varying motion blur. IEEE International 
Conference on Image Processing, Austin, TX, USA, IEEE, Los 
Alamitos, CA, USA. 
Patti, A. J., M. I. Sezan, et al. 1997. "Superresolution video 
reconstruction with arbitrary sampling lattices and nonzero