Abstract

We present experiments to illustrate tracking of occluded objects using three-dimensional (3D) integral imaging (II). Tracking of heavily occluded objects by conventional two-dimensional (2D) image processing may be difficult owing to the superposition of occlusion noise and object details. The effects of occlusion are remedied by 3D computational II reconstruction. We use a summation of absolute difference (SAD) algorithm between pixels of consecutive frames of a moving object for 3D tracking. SAD algorithm is not robust for 2D images of occluded objects; 3D computation reconstruction of the scene allows implementation of a SAD algorithm by reducing occlusion effects. Experimental results demonstrate 3D tracking of occluded objects. To the best of our knowledge, this is the first report on 3D tracking of objects using II.

© 2008 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |
  • |

  1. G. Lippmann, C. R. Acad. Bulg. Sci. 146, 446 (1908).
  2. J.-S. Jang and B. Javidi, Opt. Lett. 27, 1144 (2002).
    [CrossRef]
  3. M. Levoy, IEEE Comp. Mag. 39, 46 (2006).
  4. S.-H. Hong, J.-S. Jang, and B. Javidi, Opt. Express 12, 483 (2004).
    [CrossRef] [PubMed]
  5. H. Arimoto and B. Javidi, Opt. Lett. 26, 157 (2001).
    [CrossRef]
  6. H. Hoshino, F. Okano, H. Isono, and I. Yuyama, J. Opt. Soc. Am. A 15, 2059 (1998).
    [CrossRef]
  7. B. Javidi, S.-H. Hong, and O. Matoba, Appl. Opt. 45, 2986 (2006).
    [CrossRef] [PubMed]
  8. S. Vassiliadis, E. A. Hakkennes, J. S. S. M. Wong, and G. G. Pechanek, IEEE Comp. Soc. 2, 559 (1998).

2006 (2)

2004 (1)

2002 (1)

2001 (1)

1998 (2)

H. Hoshino, F. Okano, H. Isono, and I. Yuyama, J. Opt. Soc. Am. A 15, 2059 (1998).
[CrossRef]

S. Vassiliadis, E. A. Hakkennes, J. S. S. M. Wong, and G. G. Pechanek, IEEE Comp. Soc. 2, 559 (1998).

1908 (1)

G. Lippmann, C. R. Acad. Bulg. Sci. 146, 446 (1908).

Arimoto, H.

Hakkennes, E. A.

S. Vassiliadis, E. A. Hakkennes, J. S. S. M. Wong, and G. G. Pechanek, IEEE Comp. Soc. 2, 559 (1998).

Hong, S.-H.

Hoshino, H.

Isono, H.

Jang, J.-S.

Javidi, B.

Levoy, M.

M. Levoy, IEEE Comp. Mag. 39, 46 (2006).

Lippmann, G.

G. Lippmann, C. R. Acad. Bulg. Sci. 146, 446 (1908).

Matoba, O.

Okano, F.

Pechanek, G. G.

S. Vassiliadis, E. A. Hakkennes, J. S. S. M. Wong, and G. G. Pechanek, IEEE Comp. Soc. 2, 559 (1998).

Vassiliadis, S.

S. Vassiliadis, E. A. Hakkennes, J. S. S. M. Wong, and G. G. Pechanek, IEEE Comp. Soc. 2, 559 (1998).

Wong, J. S. S. M.

S. Vassiliadis, E. A. Hakkennes, J. S. S. M. Wong, and G. G. Pechanek, IEEE Comp. Soc. 2, 559 (1998).

Yuyama, I.

Appl. Opt. (1)

C. R. Acad. Bulg. Sci. (1)

G. Lippmann, C. R. Acad. Bulg. Sci. 146, 446 (1908).

IEEE Comp. Mag. (1)

M. Levoy, IEEE Comp. Mag. 39, 46 (2006).

IEEE Comp. Soc. (1)

S. Vassiliadis, E. A. Hakkennes, J. S. S. M. Wong, and G. G. Pechanek, IEEE Comp. Soc. 2, 559 (1998).

J. Opt. Soc. Am. A (1)

Opt. Express (1)

Opt. Lett. (2)

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 (4)

Fig. 1
Fig. 1

(a) Pickup process, (b) 3D reconstruction process, and (c) computational reconstruction process of occluded object by integral imaging.

Fig. 2
Fig. 2

3D reconstruction results with occluded object. (a) Occluded object (car behind tree branches) is located at z = 1150 mm from the sensor; (b) elemental images; (c)–(f) 3D reconstructed image sequence at z = 560 , z = 850 , z = 1150 , and z = 1320 mm , respectively. (c) Reconstructed plane of occlusion (tree branches).

Fig. 3
Fig. 3

3D moving path of occluded object. Numbers 1–15 illustrate the sequence of the moving object (see text for details).

Fig. 4
Fig. 4

3D tracking results: (a) first frame, (b) second frame, (c) third frame, (d) seventh frame, (e) eleventh frame, and (f) thirteenth frame.

Equations (8)

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

M [ e i j ( x , y ) ] = E i j ( r x × ( i 1 ) + x , r y × ( j 1 ) + y )
for x = 1 , 2 , , r x ; y = 1 , 2 , , r y ;
x = 1 2 , , round ( M × r x ) ;
y = 1 , 2 , , round ( M × r y ) ,
R ( h , v ) = 1 N s ( h , v ) i = 1 N x j = 1 N y E i j
for h = 1 , 2 , , round ( r x × ( M + N x 1 ) ) ;
v = 1 , 2 , , round ( r y × ( M + N y 1 ) ) ,
SAD ( x , y ) = p = 0 t x 1 q = 0 t y 1 R n ( x + p , y + q ) T n 1 ( p , q ) ,

Metrics