Abstract

An improved scene-adaptive nonuniformity correction (NUC) algorithm for infrared focal plane arrays (IRFPAs) is proposed. This method simultaneously estimates the infrared detectors’ parameters and eliminates the nonuniformity causing fixed pattern noise (FPN) by using a neural network (NN) approach. In the learning process of neuron parameter estimation, the traditional LMS algorithm is substituted with the newly presented variable step size (VSS) normalized least-mean square (NLMS) based adaptive filtering algorithm, which yields faster convergence, smaller misadjustment, and lower computational cost. In addition, a new NN structure is designed to estimate the desired target value, which promotes the calibration precision considerably. The proposed NUC method reaches high correction performance, which is validated by the experimental results quantitatively tested with a simulative testing sequence and a real infrared image sequence.

© 2008 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |
  • |

  1. D. Scribner, K. Sarkady, and M. Kruer, “Adaptive nonuniformity correction for IR focal-plane arrays using neural networks,” Proc. SPIE 1541, 100-109 (1991).
    [CrossRef]
  2. D. Scribner, K. Sarkady, and M. Kruer, “Adaptive retina-like preprocessing for imaging detector arrays,” in Proceeding of the IEEE Conference on Neural Network (IEEE, 1993), Vol. 3, pp. 1955-1960.
    [CrossRef]
  3. S. N. Torres, C. San Martin, D. Sbarbaro, and J. E. Pezoa, “A neural network for nonuniformity and ghosting correction of infrared image sequence,” Lect. Notes Comput. Sci. 3656, 1208-1216 (2005).
    [CrossRef]
  4. J. Kim and A. Poularikas, “Comparison of two proposed methods in adaptive noise canceling,” in Proceedings of the IEEE 35th Southeastern Symposium System Theory (IEEE, 2003), pp. 400-403.
  5. J. Kim, C.-W. Lee, J.-K Lee, J.-S.Hwang, and J.-H. Park, “Comparison of adaptive systems for noise reduction in speech performance,” in Proceedings of the IEEE 38th Southeastern Symposium on System Theory (IEEE, 2006), pp. 123-127.
  6. R. H. Kwong and E. W. Johnston, “A variable step size LMS algorithm,” IEEE Trans. Signal Process. 40, 1663-1642(1992).
    [CrossRef]
  7. D. G. Manoakis, V. K. Ingle, and S. M. Kogon, Statistical and Adaptive Signal Processing (McGraw-Hill, 2000).
  8. M. M. Hayat, S. N. Torres, E. Armstrong, S. C. Cain, and B. Yasuda, “Statistical algorithm for nonuniformity correction in focal-plane arrays,” Appl. Opt. 38, 772-780 (1999).
    [CrossRef]
  9. K. Benkrid and S. Belkacemi, “Design and implementation of a 2D convolution core for video applications on FPGAs,” in Proceedings of the IEEE International Conference on Digital and Computational Video (IEEE, 2002), pp. 85-92.
  10. S. Perri, M. Lanuzza, P. Corsonello, and G. Cocorullo, “A high-performance fully reconfigurable FPGA-based 2D convolution processor,” MicroProcess. Microsyst. 29, 381-391 (2005).
    [CrossRef]

2005 (2)

S. N. Torres, C. San Martin, D. Sbarbaro, and J. E. Pezoa, “A neural network for nonuniformity and ghosting correction of infrared image sequence,” Lect. Notes Comput. Sci. 3656, 1208-1216 (2005).
[CrossRef]

S. Perri, M. Lanuzza, P. Corsonello, and G. Cocorullo, “A high-performance fully reconfigurable FPGA-based 2D convolution processor,” MicroProcess. Microsyst. 29, 381-391 (2005).
[CrossRef]

1999 (1)

1992 (1)

R. H. Kwong and E. W. Johnston, “A variable step size LMS algorithm,” IEEE Trans. Signal Process. 40, 1663-1642(1992).
[CrossRef]

1991 (1)

D. Scribner, K. Sarkady, and M. Kruer, “Adaptive nonuniformity correction for IR focal-plane arrays using neural networks,” Proc. SPIE 1541, 100-109 (1991).
[CrossRef]

Armstrong, E.

Belkacemi, S.

K. Benkrid and S. Belkacemi, “Design and implementation of a 2D convolution core for video applications on FPGAs,” in Proceedings of the IEEE International Conference on Digital and Computational Video (IEEE, 2002), pp. 85-92.

Benkrid, K.

K. Benkrid and S. Belkacemi, “Design and implementation of a 2D convolution core for video applications on FPGAs,” in Proceedings of the IEEE International Conference on Digital and Computational Video (IEEE, 2002), pp. 85-92.

Cain, S. C.

Cocorullo, G.

S. Perri, M. Lanuzza, P. Corsonello, and G. Cocorullo, “A high-performance fully reconfigurable FPGA-based 2D convolution processor,” MicroProcess. Microsyst. 29, 381-391 (2005).
[CrossRef]

Corsonello, P.

S. Perri, M. Lanuzza, P. Corsonello, and G. Cocorullo, “A high-performance fully reconfigurable FPGA-based 2D convolution processor,” MicroProcess. Microsyst. 29, 381-391 (2005).
[CrossRef]

Hayat, M. M.

Hwang, J.-S.

J. Kim, C.-W. Lee, J.-K Lee, J.-S.Hwang, and J.-H. Park, “Comparison of adaptive systems for noise reduction in speech performance,” in Proceedings of the IEEE 38th Southeastern Symposium on System Theory (IEEE, 2006), pp. 123-127.

Ingle, V. K.

D. G. Manoakis, V. K. Ingle, and S. M. Kogon, Statistical and Adaptive Signal Processing (McGraw-Hill, 2000).

Johnston, E. W.

R. H. Kwong and E. W. Johnston, “A variable step size LMS algorithm,” IEEE Trans. Signal Process. 40, 1663-1642(1992).
[CrossRef]

Kim, J.

J. Kim, C.-W. Lee, J.-K Lee, J.-S.Hwang, and J.-H. Park, “Comparison of adaptive systems for noise reduction in speech performance,” in Proceedings of the IEEE 38th Southeastern Symposium on System Theory (IEEE, 2006), pp. 123-127.

J. Kim and A. Poularikas, “Comparison of two proposed methods in adaptive noise canceling,” in Proceedings of the IEEE 35th Southeastern Symposium System Theory (IEEE, 2003), pp. 400-403.

Kogon, S. M.

D. G. Manoakis, V. K. Ingle, and S. M. Kogon, Statistical and Adaptive Signal Processing (McGraw-Hill, 2000).

Kruer, M.

D. Scribner, K. Sarkady, and M. Kruer, “Adaptive nonuniformity correction for IR focal-plane arrays using neural networks,” Proc. SPIE 1541, 100-109 (1991).
[CrossRef]

D. Scribner, K. Sarkady, and M. Kruer, “Adaptive retina-like preprocessing for imaging detector arrays,” in Proceeding of the IEEE Conference on Neural Network (IEEE, 1993), Vol. 3, pp. 1955-1960.
[CrossRef]

Kwong, R. H.

R. H. Kwong and E. W. Johnston, “A variable step size LMS algorithm,” IEEE Trans. Signal Process. 40, 1663-1642(1992).
[CrossRef]

Lanuzza, M.

S. Perri, M. Lanuzza, P. Corsonello, and G. Cocorullo, “A high-performance fully reconfigurable FPGA-based 2D convolution processor,” MicroProcess. Microsyst. 29, 381-391 (2005).
[CrossRef]

Lee, C.-W.

J. Kim, C.-W. Lee, J.-K Lee, J.-S.Hwang, and J.-H. Park, “Comparison of adaptive systems for noise reduction in speech performance,” in Proceedings of the IEEE 38th Southeastern Symposium on System Theory (IEEE, 2006), pp. 123-127.

Lee, J.-K

J. Kim, C.-W. Lee, J.-K Lee, J.-S.Hwang, and J.-H. Park, “Comparison of adaptive systems for noise reduction in speech performance,” in Proceedings of the IEEE 38th Southeastern Symposium on System Theory (IEEE, 2006), pp. 123-127.

Manoakis, D. G.

D. G. Manoakis, V. K. Ingle, and S. M. Kogon, Statistical and Adaptive Signal Processing (McGraw-Hill, 2000).

Park, J.-H.

J. Kim, C.-W. Lee, J.-K Lee, J.-S.Hwang, and J.-H. Park, “Comparison of adaptive systems for noise reduction in speech performance,” in Proceedings of the IEEE 38th Southeastern Symposium on System Theory (IEEE, 2006), pp. 123-127.

Perri, S.

S. Perri, M. Lanuzza, P. Corsonello, and G. Cocorullo, “A high-performance fully reconfigurable FPGA-based 2D convolution processor,” MicroProcess. Microsyst. 29, 381-391 (2005).
[CrossRef]

Pezoa, J. E.

S. N. Torres, C. San Martin, D. Sbarbaro, and J. E. Pezoa, “A neural network for nonuniformity and ghosting correction of infrared image sequence,” Lect. Notes Comput. Sci. 3656, 1208-1216 (2005).
[CrossRef]

Poularikas, A.

J. Kim and A. Poularikas, “Comparison of two proposed methods in adaptive noise canceling,” in Proceedings of the IEEE 35th Southeastern Symposium System Theory (IEEE, 2003), pp. 400-403.

San Martin, C.

S. N. Torres, C. San Martin, D. Sbarbaro, and J. E. Pezoa, “A neural network for nonuniformity and ghosting correction of infrared image sequence,” Lect. Notes Comput. Sci. 3656, 1208-1216 (2005).
[CrossRef]

Sarkady, K.

D. Scribner, K. Sarkady, and M. Kruer, “Adaptive nonuniformity correction for IR focal-plane arrays using neural networks,” Proc. SPIE 1541, 100-109 (1991).
[CrossRef]

D. Scribner, K. Sarkady, and M. Kruer, “Adaptive retina-like preprocessing for imaging detector arrays,” in Proceeding of the IEEE Conference on Neural Network (IEEE, 1993), Vol. 3, pp. 1955-1960.
[CrossRef]

Sbarbaro, D.

S. N. Torres, C. San Martin, D. Sbarbaro, and J. E. Pezoa, “A neural network for nonuniformity and ghosting correction of infrared image sequence,” Lect. Notes Comput. Sci. 3656, 1208-1216 (2005).
[CrossRef]

Scribner, D.

D. Scribner, K. Sarkady, and M. Kruer, “Adaptive nonuniformity correction for IR focal-plane arrays using neural networks,” Proc. SPIE 1541, 100-109 (1991).
[CrossRef]

D. Scribner, K. Sarkady, and M. Kruer, “Adaptive retina-like preprocessing for imaging detector arrays,” in Proceeding of the IEEE Conference on Neural Network (IEEE, 1993), Vol. 3, pp. 1955-1960.
[CrossRef]

Torres, S. N.

S. N. Torres, C. San Martin, D. Sbarbaro, and J. E. Pezoa, “A neural network for nonuniformity and ghosting correction of infrared image sequence,” Lect. Notes Comput. Sci. 3656, 1208-1216 (2005).
[CrossRef]

M. M. Hayat, S. N. Torres, E. Armstrong, S. C. Cain, and B. Yasuda, “Statistical algorithm for nonuniformity correction in focal-plane arrays,” Appl. Opt. 38, 772-780 (1999).
[CrossRef]

Yasuda, B.

Appl. Opt. (1)

IEEE Trans. Signal Process. (1)

R. H. Kwong and E. W. Johnston, “A variable step size LMS algorithm,” IEEE Trans. Signal Process. 40, 1663-1642(1992).
[CrossRef]

Lect. Notes Comput. Sci. (1)

S. N. Torres, C. San Martin, D. Sbarbaro, and J. E. Pezoa, “A neural network for nonuniformity and ghosting correction of infrared image sequence,” Lect. Notes Comput. Sci. 3656, 1208-1216 (2005).
[CrossRef]

MicroProcess. Microsyst. (1)

S. Perri, M. Lanuzza, P. Corsonello, and G. Cocorullo, “A high-performance fully reconfigurable FPGA-based 2D convolution processor,” MicroProcess. Microsyst. 29, 381-391 (2005).
[CrossRef]

Proc. SPIE (1)

D. Scribner, K. Sarkady, and M. Kruer, “Adaptive nonuniformity correction for IR focal-plane arrays using neural networks,” Proc. SPIE 1541, 100-109 (1991).
[CrossRef]

Other (5)

D. Scribner, K. Sarkady, and M. Kruer, “Adaptive retina-like preprocessing for imaging detector arrays,” in Proceeding of the IEEE Conference on Neural Network (IEEE, 1993), Vol. 3, pp. 1955-1960.
[CrossRef]

J. Kim and A. Poularikas, “Comparison of two proposed methods in adaptive noise canceling,” in Proceedings of the IEEE 35th Southeastern Symposium System Theory (IEEE, 2003), pp. 400-403.

J. Kim, C.-W. Lee, J.-K Lee, J.-S.Hwang, and J.-H. Park, “Comparison of adaptive systems for noise reduction in speech performance,” in Proceedings of the IEEE 38th Southeastern Symposium on System Theory (IEEE, 2006), pp. 123-127.

D. G. Manoakis, V. K. Ingle, and S. M. Kogon, Statistical and Adaptive Signal Processing (McGraw-Hill, 2000).

K. Benkrid and S. Belkacemi, “Design and implementation of a 2D convolution core for video applications on FPGAs,” in Proceedings of the IEEE International Conference on Digital and Computational Video (IEEE, 2002), pp. 85-92.

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

Scheme of the proposed scene-adaptive nonuniformity correction method.

Fig. 2
Fig. 2

RMSE between the artificially corrupted sequence and its corrected sequence. Dashed–dotted line, corrupted data; dotted line, Scribner’s method; line with circles, Torres’s enhanced method; solid line, our proposed method.

Fig. 3
Fig. 3

Visual effect of different NUC methods, in which (a) is the 725th frame in the real infrared sequence and (b), (c) and (d) are the correction results of Scribner’s method, Torres’s method and our proposed method after 15 iterations, respectively. The white boxes mark partial obvious lattice shaped noise jamming in images.

Fig. 4
Fig. 4

Run time comparison of different methods.

Tables (2)

Tables Icon

Table 1 Performance Parameters ρ for Artificial Data

Tables Icon

Table 2 Performance Parameter ρ for Real Infrared Data

Equations (14)

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

Y ^ i j ( n ) = W i j T ( n ) · X i j ( n ) ,
E i j ( n ) = T i j ( n ) Y ^ i j ( n ) ,
T i j ( n ) = 1 4 L ( L + 1 ) u [ i L , i ) ( i , i + L ] v [ j L , j ) ( j , j + L ] X u v ( n ) .
W i j ( n + 1 ) = W i j ( n ) + 2 μ · E i j ( n ) · X i j ( n ) .
g ^ i j ( n + 1 ) = g ^ i j ( n ) + η · E i j ( n ) · x i j ( n ) ,
o ^ i j ( n + 1 ) = o ^ i j ( n ) + η · E i j ( n ) ,
T i j ( n ) = u = i L i + L v = j L j + L K u v ( n ) · X u v ( n ) u = i L i + L v = j L j + L K u v ( n ) ,
K u v ( n ) = exp ( d 2 ( [ u , v ] , [ i , j ] ) 2 σ d 2 ) ,
W i j ( n + 1 ) = W i j ( n ) + η ˜ n X i j T ( n ) · X i j ( n ) + c · E i j ( n ) · X i j ( n ) ,
g ^ i j ( n + 1 ) = g ^ i j ( n ) + η ˜ n x i j 2 ( n ) + c · E i j ( n ) · x i j ( n ) ,
o ^ i j ( n + 1 ) = o ^ i j ( n ) + η ˜ n x i j 2 ( n ) + c · E i j ( n ) ,
η ˜ k + 1 = { η max , if     η k + 1 > η max η min , if     η k + 1 < η min η k + 1 , otherwise ,
RMSE ( n ) = [ 1 M N i = 1 M j = 1 N ( Y ^ i j ( n ) Y i j ( n ) ) 2 ] 1 / 2 .
ρ = h 1 * f 1 + h 2 * f 1 f 1 ,

Metrics