Abstract

We present an adaptive autofocusing scheme. In this scheme, the focus measure is updated with focus tuning. To achieve this, we construct the focus measure by using image moments and develop an adaptive focus-tuning strategy to estimate the measure in closed loop. It is shown that the adaptive updating of the focus measure enables us to overcome the dependence of autofocusing on the image contents. Such an adaptive closed-loop focusing operation also effectively suppresses both the effect of the noise in optical imaging and the effect of time delay due to image processing time. Therefore a high accuracy of autofocusing is guaranteed. The effectiveness of the proposed scheme is demonstrated by simulations and experiments.

© 2005 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |
  • |

  1. A. Pentland, “New sense for depth of field,” IEEE Trans. Pattern Anal. Mach. Intell. PAMI-9, 523–531 (1987).
    [CrossRef]
  2. L.-M. Koh, H.-O. Lim, “Computer vision techniques for IC wire-bond height inspection,” in Automatic Inspection and Novel Instrumentation, A. T. S. Ho, S. Rao, L. M. Cheng, eds., Proc. SPIE3185, 11–21 (1997).
    [CrossRef]
  3. R. A. Jarvis, “Focus optimization criteria for computer image processing,” Microscope 24, 163–180 (1976).
  4. E. Krotkov, E. Paul, Active Computer Vision by Cooperative Focus and Stereo (Springer-Verlag, New York, 1989).
  5. M. Chargi, A. Nyeck, A. Tosser, “Focusing criterion,” Electron. Lett. 27, 1233–1235 (1991).
    [CrossRef]
  6. M. Subbarao, T. Choi, “Focusing techniques,” Opt. Eng. 32, 2824–2836 (1993).
    [CrossRef]
  7. M. Subbarao, J. Tyan, “Selecting the optimal focus measure for auto-focusing and depth-from-focus,” IEEE Trans. Pattern Anal. Mach. Intell. 20, 864–870 (1998).
    [CrossRef]
  8. S. Nayar, Y. Nakagawa, “Shape from focus,” IEEE Trans. Pattern Anal. Mach. Intell. 16, 824–831 (1994).
    [CrossRef]
  9. Y. Xiong, S. A. Shafer, “Moment filters for high precision computational of focus and stereo,” in Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IEEE Press, Piscataway, N.J., 1995), Vol. 3, pp. 108–113.
  10. S. J. J. Widjaja, “Use of wavelet analysis for improving autofocusing capability,” Opt. Commun. 151(1–3), 12–14 (1998).
    [CrossRef]
  11. Y. Zhang, Y. Zhang, C. Wen, “A new focus measure method using moments,” Int. J. Image Vision Comput. 18, 959–965 (2000).
    [CrossRef]
  12. G. C. Goodwin, K. S. Sin, Adaptive Filtering: Prediction and Control (Prentice Hall, Englewood Cliffs, N.J., 1984).

2000 (1)

Y. Zhang, Y. Zhang, C. Wen, “A new focus measure method using moments,” Int. J. Image Vision Comput. 18, 959–965 (2000).
[CrossRef]

1998 (2)

S. J. J. Widjaja, “Use of wavelet analysis for improving autofocusing capability,” Opt. Commun. 151(1–3), 12–14 (1998).
[CrossRef]

M. Subbarao, J. Tyan, “Selecting the optimal focus measure for auto-focusing and depth-from-focus,” IEEE Trans. Pattern Anal. Mach. Intell. 20, 864–870 (1998).
[CrossRef]

1994 (1)

S. Nayar, Y. Nakagawa, “Shape from focus,” IEEE Trans. Pattern Anal. Mach. Intell. 16, 824–831 (1994).
[CrossRef]

1993 (1)

M. Subbarao, T. Choi, “Focusing techniques,” Opt. Eng. 32, 2824–2836 (1993).
[CrossRef]

1991 (1)

M. Chargi, A. Nyeck, A. Tosser, “Focusing criterion,” Electron. Lett. 27, 1233–1235 (1991).
[CrossRef]

1987 (1)

A. Pentland, “New sense for depth of field,” IEEE Trans. Pattern Anal. Mach. Intell. PAMI-9, 523–531 (1987).
[CrossRef]

1976 (1)

R. A. Jarvis, “Focus optimization criteria for computer image processing,” Microscope 24, 163–180 (1976).

Chargi, M.

M. Chargi, A. Nyeck, A. Tosser, “Focusing criterion,” Electron. Lett. 27, 1233–1235 (1991).
[CrossRef]

Choi, T.

M. Subbarao, T. Choi, “Focusing techniques,” Opt. Eng. 32, 2824–2836 (1993).
[CrossRef]

Goodwin, G. C.

G. C. Goodwin, K. S. Sin, Adaptive Filtering: Prediction and Control (Prentice Hall, Englewood Cliffs, N.J., 1984).

Jarvis, R. A.

R. A. Jarvis, “Focus optimization criteria for computer image processing,” Microscope 24, 163–180 (1976).

Koh, L.-M.

L.-M. Koh, H.-O. Lim, “Computer vision techniques for IC wire-bond height inspection,” in Automatic Inspection and Novel Instrumentation, A. T. S. Ho, S. Rao, L. M. Cheng, eds., Proc. SPIE3185, 11–21 (1997).
[CrossRef]

Krotkov, E.

E. Krotkov, E. Paul, Active Computer Vision by Cooperative Focus and Stereo (Springer-Verlag, New York, 1989).

Lim, H.-O.

L.-M. Koh, H.-O. Lim, “Computer vision techniques for IC wire-bond height inspection,” in Automatic Inspection and Novel Instrumentation, A. T. S. Ho, S. Rao, L. M. Cheng, eds., Proc. SPIE3185, 11–21 (1997).
[CrossRef]

Nakagawa, Y.

S. Nayar, Y. Nakagawa, “Shape from focus,” IEEE Trans. Pattern Anal. Mach. Intell. 16, 824–831 (1994).
[CrossRef]

Nayar, S.

S. Nayar, Y. Nakagawa, “Shape from focus,” IEEE Trans. Pattern Anal. Mach. Intell. 16, 824–831 (1994).
[CrossRef]

Nyeck, A.

M. Chargi, A. Nyeck, A. Tosser, “Focusing criterion,” Electron. Lett. 27, 1233–1235 (1991).
[CrossRef]

Paul, E.

E. Krotkov, E. Paul, Active Computer Vision by Cooperative Focus and Stereo (Springer-Verlag, New York, 1989).

Pentland, A.

A. Pentland, “New sense for depth of field,” IEEE Trans. Pattern Anal. Mach. Intell. PAMI-9, 523–531 (1987).
[CrossRef]

Shafer, S. A.

Y. Xiong, S. A. Shafer, “Moment filters for high precision computational of focus and stereo,” in Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IEEE Press, Piscataway, N.J., 1995), Vol. 3, pp. 108–113.

Sin, K. S.

G. C. Goodwin, K. S. Sin, Adaptive Filtering: Prediction and Control (Prentice Hall, Englewood Cliffs, N.J., 1984).

Subbarao, M.

M. Subbarao, J. Tyan, “Selecting the optimal focus measure for auto-focusing and depth-from-focus,” IEEE Trans. Pattern Anal. Mach. Intell. 20, 864–870 (1998).
[CrossRef]

M. Subbarao, T. Choi, “Focusing techniques,” Opt. Eng. 32, 2824–2836 (1993).
[CrossRef]

Tosser, A.

M. Chargi, A. Nyeck, A. Tosser, “Focusing criterion,” Electron. Lett. 27, 1233–1235 (1991).
[CrossRef]

Tyan, J.

M. Subbarao, J. Tyan, “Selecting the optimal focus measure for auto-focusing and depth-from-focus,” IEEE Trans. Pattern Anal. Mach. Intell. 20, 864–870 (1998).
[CrossRef]

Wen, C.

Y. Zhang, Y. Zhang, C. Wen, “A new focus measure method using moments,” Int. J. Image Vision Comput. 18, 959–965 (2000).
[CrossRef]

Widjaja, S. J. J.

S. J. J. Widjaja, “Use of wavelet analysis for improving autofocusing capability,” Opt. Commun. 151(1–3), 12–14 (1998).
[CrossRef]

Xiong, Y.

Y. Xiong, S. A. Shafer, “Moment filters for high precision computational of focus and stereo,” in Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IEEE Press, Piscataway, N.J., 1995), Vol. 3, pp. 108–113.

Zhang, Y.

Y. Zhang, Y. Zhang, C. Wen, “A new focus measure method using moments,” Int. J. Image Vision Comput. 18, 959–965 (2000).
[CrossRef]

Y. Zhang, Y. Zhang, C. Wen, “A new focus measure method using moments,” Int. J. Image Vision Comput. 18, 959–965 (2000).
[CrossRef]

Electron. Lett. (1)

M. Chargi, A. Nyeck, A. Tosser, “Focusing criterion,” Electron. Lett. 27, 1233–1235 (1991).
[CrossRef]

IEEE Trans. Pattern Anal. Mach. Intell. (3)

M. Subbarao, J. Tyan, “Selecting the optimal focus measure for auto-focusing and depth-from-focus,” IEEE Trans. Pattern Anal. Mach. Intell. 20, 864–870 (1998).
[CrossRef]

S. Nayar, Y. Nakagawa, “Shape from focus,” IEEE Trans. Pattern Anal. Mach. Intell. 16, 824–831 (1994).
[CrossRef]

A. Pentland, “New sense for depth of field,” IEEE Trans. Pattern Anal. Mach. Intell. PAMI-9, 523–531 (1987).
[CrossRef]

Int. J. Image Vision Comput. (1)

Y. Zhang, Y. Zhang, C. Wen, “A new focus measure method using moments,” Int. J. Image Vision Comput. 18, 959–965 (2000).
[CrossRef]

Microscope (1)

R. A. Jarvis, “Focus optimization criteria for computer image processing,” Microscope 24, 163–180 (1976).

Opt. Commun. (1)

S. J. J. Widjaja, “Use of wavelet analysis for improving autofocusing capability,” Opt. Commun. 151(1–3), 12–14 (1998).
[CrossRef]

Opt. Eng. (1)

M. Subbarao, T. Choi, “Focusing techniques,” Opt. Eng. 32, 2824–2836 (1993).
[CrossRef]

Other (4)

Y. Xiong, S. A. Shafer, “Moment filters for high precision computational of focus and stereo,” in Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IEEE Press, Piscataway, N.J., 1995), Vol. 3, pp. 108–113.

E. Krotkov, E. Paul, Active Computer Vision by Cooperative Focus and Stereo (Springer-Verlag, New York, 1989).

L.-M. Koh, H.-O. Lim, “Computer vision techniques for IC wire-bond height inspection,” in Automatic Inspection and Novel Instrumentation, A. T. S. Ho, S. Rao, L. M. Cheng, eds., Proc. SPIE3185, 11–21 (1997).
[CrossRef]

G. C. Goodwin, K. S. Sin, Adaptive Filtering: Prediction and Control (Prentice Hall, Englewood Cliffs, N.J., 1984).

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 (11)

Fig. 1
Fig. 1

Schematic of an autofocusing system.

Fig. 2
Fig. 2

Schematic of image formation.

Fig. 3
Fig. 3

Schematic of an autofocusing system subject to a self-tuning controller.

Fig. 4
Fig. 4

Performance of the proposed focusing scheme in the presence of different image contents: (a) synthetic white box, (b) portrait of Lena, (c) military vehicle, (d) microscopic image of fungus, (e) blurred image of the synthetic box ( σ = 3.0 ) , (f) blurred image of Lena’s portrait ( σ = 3.0 ) , (g) blurred image of the military vehicle ( σ = 3.0 ) , (h) blurred microscopy image of fungus ( σ = 3.0 ) .

Fig. 5
Fig. 5

Moment focus measures of the out-of-focus images in Fig. 4: (a) synthetic box, (b) Lena’s portrait, (c) military vehicle, (d) microscopy image of fungus.

Fig. 6
Fig. 6

Performance of the adaptive focusing scheme with respect to the window size: (a) window size 256 × 256 , (b) window size 128 × 128 , (c) window size 64 × 64 , (d) window size 256 × 256 , (e) window size 128 × 128 , (f) window size 64 × 64 .

Fig. 7
Fig. 7

Synthetic image to be focused and its associated moments: (a) in-focus image, (b) in-focus image with noise.

Fig. 8
Fig. 8

Passive focus measures: (a) ideal case without noise, (b) case with Gaussian noise ( SNR = 20   dB ) .

Fig. 9
Fig. 9

Results of the proposed active focusing scheme: (a) ideal case without noise, (b) case with Gaussian noise ( SNR = 20   dB ) .

Fig. 10
Fig. 10

Experimental results of the autofocusing scheme: (a) focused image, (b) out-of-focus image.

Fig. 11
Fig. 11

Experimental results of the autofocusing scheme.

Equations (39)

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

x ˙ ac ( t ) = Ax ac ( t ) + Bu ( t ) ,
z ( t ) = Cx ac ( t ) + Du ( t ) ,
1 F = 1 z 0 + 1 v .
R ( D / 2 ) = | v - v | v .
1 F = 1 v + 1 z .
R = Dv 2 F - D 2 - vD 2 z .
g ( x ,   y ) = f ( x ,   y )     *     h ( x ,   y ) + n ( x ,   y ) ,
- - h ( x ,   y ) d x d y = 1 .
1 2 π σ 2 exp - x 2 + y 2 2 σ 2 ,
σ = kD 2 v F - 1 - v z ,
M = μ 02 ( g ) = μ 00 ( f ) σ 2 + μ 02 ( f ) ,
μ pg ( f ) - - ( x - x c ) p ( y - y c ) q d x d y ,
x c = - - xf ( x ,   y ) d x d y - - f ( x ,   y ) d x d y ,
y c = - - yf ( x ,   y ) d x d y - - f ( x ,   y ) d x d y .
M = μ 00 ( f ) k 2 D 2 4 v F - 1 - v z 2 + μ 02 ( f ) .
M = a   1 z 2 - b   1 z + c ,
A ( q - 1 ) z ( t ) = B ( q - 1 ) u ( t ) ,
A ( q - 1 ) = 1 - a 1 q - 1 - a 2 q - 2 - - a n q - n ,
B ( q - 1 ) = q - k ( b 0 + b 1 q - 1 + b 2 q - 2 + + b m q - m ) = q - k B ( q - 1 ) ,
z ( t + k ) = α ( q - 1 ) z ( t ) + β ( q - 1 ) u ( t ) ,
F ( q - 1 ) A ( q - 1 ) + q - d G ( q - 1 ) = 1 .
w ( t + k ) = a   1 z ( t + k ) 2 - b   1 z ( t + k ) + c .
J u ( t ) = 0 .
u ( t ) = 1 β 0 - α ( q - 1 ) z ( t ) - β ( q - 1 ) u ( t ) - b 2 a ,
ϕ L ( t - 1 ) = [ z ( t - 1 ) ,   z ( t - 2 ) , ,   z ( t - n ) ;   u ( t ) ,   u ( t - 1 ) , ,   u ( t - m ) ] T ,
θ L = [ a 1 ,   a 2 , ,   a n ;   b 1 ,   b 2 , ,   b m ] T ,
ϕ NL ( t - 1 ) = 1 z ( t - 1 ) 2 ,   - 1 z ( t - 1 ) ,   1 T ,
θ NL = [ a ,   b ,   c ] T .
z ( t ) = ϕ L ( t - 1 ) T θ L ,
w ( t ) = ϕ NL ( t ) T θ NL .
θ ^ L ( t ) = θ ^ L ( t - 1 ) + P L ( t - 1 ) ϕ L ( t - 1 ) 1 + ϕ L ( t - 1 ) T P L ( t - 1 ) ϕ L ( t - 1 )   [ z ( t ) - ϕ L ( t - 1 ) T θ ^ L ( t - 1 ) ] ,
θ ^ NL ( t ) = θ ^ NL ( t - 1 ) + P NL ( t - 1 ) ϕ NL ( t - 1 ) 1 + ϕ NL ( t - 1 ) T P NL ( t - 1 ) ϕ NL ( t - 1 )   × [ w ( t ) - ϕ NL ( t ) T θ ^ NL ( t - 1 ) ] ,
P L ( t ) = P L ( t - 1 ) - P L ( t - 1 ) ϕ L ( t - 1 ) ϕ L ( t - 1 ) T P L ( t ) 1 + ϕ L ( t - 1 ) T P L ( t - 1 ) ϕ L ( t - 1 ) ,
P NL ( t ) = P NL ( t - 1 )
- P NL ( t - 1 ) ϕ NL ( t - 1 ) ϕ NL ( t - 1 ) T P NL ( t - 1 ) 1 + ϕ NL ( t - 1 ) T P NL ( t - 1 ) ϕ NL ( t - 1 ) .
u ( t ) = 1 β ^ 0 - α ˆ ( q - 1 ) z ( t ) - β ^ ( q - 1 ) u ( t ) - θ ^ NL 2 ( t - 1 ) 2 θ ^ NL 1 ( t - 1 ) ,
α ˆ ( q - 1 ) = G ˆ ( q - 1 ) ,
β ˆ ( q - 1 ) = F ˆ ( q - 1 ) B ^ ( q - 1 ) ,
β ^ ( q - 1 ) = q ( β ˆ ( q - 1 ) - β ^ 0 ) .

Metrics