Abstract

One approach to flat sensor design is to use a lenslet array to form multiple subimages of a scene and then combine the subimages to recover a fully sampled image by using a superresolution algorithm. Previously, superresolution image assembly has been based on information derived from the observed scene. For lenslet arrays, we propose a new scene-independent approach based only on known imager properties in which relative subimage shifts are accurately estimated with a calibration procedure using point source imaging. Thus, the relative resolution enhancement provided by the scene-independent superresolution algorithm is impervious to changes in subimage content, contrast, sharpness, and noise.

© 2007 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |

  1. J. Tanida, T. Kumagai, K. Yamada, S. Miyatake, K. Ishida, T. Morimoto, N. Kondou, D. Miyazaki, and Y. Ichioka, Appl. Opt. 40, 1806 (2001).
    [CrossRef]
  2. K. Nitta, R. Shogenji, S. Myiatake, and J. Tanida, Appl. Opt. 45, 2893 (2006).
    [CrossRef] [PubMed]
  3. M. W. Haney, Appl. Opt. 45, 2901 (2006).
    [CrossRef] [PubMed]
  4. A. V. Kanaev, D. A. Scribner, E. F. Fleet, and J. R. Ackerman, Appl. Opt. 46, 4320 (2007).
    [CrossRef] [PubMed]
  5. S. C. Park, M. K. Park, and M. Gi Kang, IEEE Signal Process. Mag. 3, 21 (2003); and references herein.
    [CrossRef]
  6. T. Q. Pham, M. Bezuijen, L. J. Vliet, K. Schutte, and C. L. Luengo Hendriks, Proc. SPIE 5817, 133 (2005).
    [CrossRef]

2007

2006

2005

T. Q. Pham, M. Bezuijen, L. J. Vliet, K. Schutte, and C. L. Luengo Hendriks, Proc. SPIE 5817, 133 (2005).
[CrossRef]

2003

S. C. Park, M. K. Park, and M. Gi Kang, IEEE Signal Process. Mag. 3, 21 (2003); and references herein.
[CrossRef]

2001

Cited By

OSA participates in CrossRef's Cited-By Linking service. Citing articles from OSA journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (3)

Fig. 1
Fig. 1

TOMBO subimages: (a) original, (b) shot noise added ( 5 × 10 4 I ) , (c) power-law gray-scale transformed ( γ = 3 ) and 100 × dynamic range reduced, (d) interpolated, (e) shot noise added ( 5 × 10 4 I ) and interpolated, (f) power-law gray-scale transformed with 100 × dynamic range reduced and interpolated.

Fig. 2
Fig. 2

Processed images: (a) a priori calibrated 8 × superresolved original; (b) a priori calibrated 8 × superresolved, shot noise added; (c) a priori calibrated 8 × superresolved, gray-scale transformed, and dynamic range reduced; (d) 8 × registration based superresolved original; (e) 8 × registration-based superresolved, shot noise added; (f) 8 × registration-based superresolved gray-scale transformed and dynamic range reduced.

Fig. 3
Fig. 3

64 low-resolution submages of “F” are simulated with their relative global shifts constituting a rectangular grid (shown). Registration estimated global shifts are shown as dots for subimages: (a) original, (b) shot noise added, (c) gray-scale transformed and dynamic range reduced. Corresponding superresolved images are shown in the upper corners for registration-based (left) and for a priori known (right) displacements.

Equations (1)

Equations on this page are rendered with MathJax. Learn more.

[ x , y ( x s q ) 2 x , y x s p y s q x , y x s p y s q x , y ( y s q ) 2 ] [ h i v i ] = [ x , y ( s p s q ) x s q x , y ( s p s q ) y s q ] .

Metrics