Abstract

We address the problem of target segmentation in active polarimetric images, which can reveal contrasts that do not appear in standard intensity images. However, these images are perturbed by strong specklelike noise. For the purpose of segmentation we thus use statistical active contours, which are known to possess noise robustness properties. The polarimetric imagers we consider acquire two different images of the same scene so as to form a two-channel image (TCI). These two images can be combined to form the orthogonal state contrast image (OSCI), which represents the degree of polarization of the backscattered light if its coherency matrix is diagonal. We characterize the segmentation performance of the statistical active contour procedure on the TCI and on the OSCI. In particular, we show that if the illumination beam is spatially nonuniform, it is more efficient to perform the segmentation on the OSCI, which is independent of the spatial variations of the illumination.

© 2002 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |
  • |

  1. W. G. Egan, W. R. Johnson, V. S. Whitehead, “Terrestrial polarization imagery obtained from the space shuttle: characterization and interpretation,” Appl. Opt. 30, 435–442 (1991).
    [CrossRef] [PubMed]
  2. J. L. Pezzaniti, R. A. Chipman, “Mueller matrix imaging polarimetry,” Opt. Eng. 34, 1558–1568 (1995).
    [CrossRef]
  3. R. A. Chipman, “Polarization diversity active imaging,” in Image Reconstruction and Restoration II, T. J. Schulz, ed., Proc. SPIE3170, 68–73 (1997).
    [CrossRef]
  4. B. Johnson, R. Joseph, M. L. Nischan, A. Newbury, J. P. Kerekes, H. T. Barclay, B. C. Willard, J. J. Zayhowski, “Compact active hyperspectral imaging system for the detection of concealed targets,” in Detection and Remediation Technologies for Mines and Minelike Targets IV, A. C. Dubey, J. F. Harvey, J. T. Broach, R. E. Dugan, eds., Proc. SPIE3710, 144–153 (1999).
    [CrossRef]
  5. S. Breugnot, Ph. Clémenceau, “Modeling and performances of a polarization active imager at lambda = 806 nm,” in Laser Radar Technology and Applications IV, G. W. Kamerman, C. Werner, eds., Proc. SPIE3707, 449–460 (1999).
    [CrossRef]
  6. J. W. Goodman, “The speckle effect in coherent imaging,” in Statistical Optics, (Wiley, New York, 1985), pp. 347–356.
  7. O. Germain, Ph. Réfrégier, “Optimal snake-based segmentation of a random luminance target on a spatially disjoint background,” Opt. Lett. 21, 1845–1847 (1996).
    [CrossRef] [PubMed]
  8. C. Chesnaud, Ph. Réfrégier, V. Boulet, “Statistical region snake-based segmentation adapted to different physical noise models,” IEEE Trans. Pattern. Anal. Mach. Intell. 21, 1145–1157 (1999).
    [CrossRef]
  9. Ph. Réfrégier, O. Germain, T. Gaidon, “Optimal snake segmentation of target and background with independent gamma density probabilities, application to speckled and preprocessed images,” Opt. Commun. 137, 382–388 (1997).
    [CrossRef]
  10. O. Germain, Ph. Réfrégier, “Snake-based method for the segmentation of objects in multichannel images degraded by speckle,” Opt. Lett. 24, 814–816 (1999).
    [CrossRef]
  11. P. Clémenceau, S. Breugnot, L. Collot, “Polarization diversity imaging,” in Laser Radar Technology and Applications III, Proc. SPIE3380, 284–291 (1998).
    [CrossRef]
  12. J. W. Goodman, “;Laser Speckle and Related Phenomena,” in Statistical Properties of Laser Speckle Patterns, 9–75 [Springer-Verlag (Topics in Applied Physics Vol. 9), Heidelberg, 1975], pp. 9–75.
  13. F. Goudail, V. Pagé, Ph. Réfrégier, “Improving target detection with polarization diversity imaging,” in Second Symposium on Physics in Signal and Image Processing (SEE, Paris, 2001). pp. 340–345.

1999 (2)

C. Chesnaud, Ph. Réfrégier, V. Boulet, “Statistical region snake-based segmentation adapted to different physical noise models,” IEEE Trans. Pattern. Anal. Mach. Intell. 21, 1145–1157 (1999).
[CrossRef]

O. Germain, Ph. Réfrégier, “Snake-based method for the segmentation of objects in multichannel images degraded by speckle,” Opt. Lett. 24, 814–816 (1999).
[CrossRef]

1997 (1)

Ph. Réfrégier, O. Germain, T. Gaidon, “Optimal snake segmentation of target and background with independent gamma density probabilities, application to speckled and preprocessed images,” Opt. Commun. 137, 382–388 (1997).
[CrossRef]

1996 (1)

1995 (1)

J. L. Pezzaniti, R. A. Chipman, “Mueller matrix imaging polarimetry,” Opt. Eng. 34, 1558–1568 (1995).
[CrossRef]

1991 (1)

Barclay, H. T.

B. Johnson, R. Joseph, M. L. Nischan, A. Newbury, J. P. Kerekes, H. T. Barclay, B. C. Willard, J. J. Zayhowski, “Compact active hyperspectral imaging system for the detection of concealed targets,” in Detection and Remediation Technologies for Mines and Minelike Targets IV, A. C. Dubey, J. F. Harvey, J. T. Broach, R. E. Dugan, eds., Proc. SPIE3710, 144–153 (1999).
[CrossRef]

Boulet, V.

C. Chesnaud, Ph. Réfrégier, V. Boulet, “Statistical region snake-based segmentation adapted to different physical noise models,” IEEE Trans. Pattern. Anal. Mach. Intell. 21, 1145–1157 (1999).
[CrossRef]

Breugnot, S.

P. Clémenceau, S. Breugnot, L. Collot, “Polarization diversity imaging,” in Laser Radar Technology and Applications III, Proc. SPIE3380, 284–291 (1998).
[CrossRef]

S. Breugnot, Ph. Clémenceau, “Modeling and performances of a polarization active imager at lambda = 806 nm,” in Laser Radar Technology and Applications IV, G. W. Kamerman, C. Werner, eds., Proc. SPIE3707, 449–460 (1999).
[CrossRef]

Chesnaud, C.

C. Chesnaud, Ph. Réfrégier, V. Boulet, “Statistical region snake-based segmentation adapted to different physical noise models,” IEEE Trans. Pattern. Anal. Mach. Intell. 21, 1145–1157 (1999).
[CrossRef]

Chipman, R. A.

J. L. Pezzaniti, R. A. Chipman, “Mueller matrix imaging polarimetry,” Opt. Eng. 34, 1558–1568 (1995).
[CrossRef]

R. A. Chipman, “Polarization diversity active imaging,” in Image Reconstruction and Restoration II, T. J. Schulz, ed., Proc. SPIE3170, 68–73 (1997).
[CrossRef]

Clémenceau, P.

P. Clémenceau, S. Breugnot, L. Collot, “Polarization diversity imaging,” in Laser Radar Technology and Applications III, Proc. SPIE3380, 284–291 (1998).
[CrossRef]

Clémenceau, Ph.

S. Breugnot, Ph. Clémenceau, “Modeling and performances of a polarization active imager at lambda = 806 nm,” in Laser Radar Technology and Applications IV, G. W. Kamerman, C. Werner, eds., Proc. SPIE3707, 449–460 (1999).
[CrossRef]

Collot, L.

P. Clémenceau, S. Breugnot, L. Collot, “Polarization diversity imaging,” in Laser Radar Technology and Applications III, Proc. SPIE3380, 284–291 (1998).
[CrossRef]

Egan, W. G.

Gaidon, T.

Ph. Réfrégier, O. Germain, T. Gaidon, “Optimal snake segmentation of target and background with independent gamma density probabilities, application to speckled and preprocessed images,” Opt. Commun. 137, 382–388 (1997).
[CrossRef]

Germain, O.

Goodman, J. W.

J. W. Goodman, “The speckle effect in coherent imaging,” in Statistical Optics, (Wiley, New York, 1985), pp. 347–356.

J. W. Goodman, “;Laser Speckle and Related Phenomena,” in Statistical Properties of Laser Speckle Patterns, 9–75 [Springer-Verlag (Topics in Applied Physics Vol. 9), Heidelberg, 1975], pp. 9–75.

Goudail, F.

F. Goudail, V. Pagé, Ph. Réfrégier, “Improving target detection with polarization diversity imaging,” in Second Symposium on Physics in Signal and Image Processing (SEE, Paris, 2001). pp. 340–345.

Johnson, B.

B. Johnson, R. Joseph, M. L. Nischan, A. Newbury, J. P. Kerekes, H. T. Barclay, B. C. Willard, J. J. Zayhowski, “Compact active hyperspectral imaging system for the detection of concealed targets,” in Detection and Remediation Technologies for Mines and Minelike Targets IV, A. C. Dubey, J. F. Harvey, J. T. Broach, R. E. Dugan, eds., Proc. SPIE3710, 144–153 (1999).
[CrossRef]

Johnson, W. R.

Joseph, R.

B. Johnson, R. Joseph, M. L. Nischan, A. Newbury, J. P. Kerekes, H. T. Barclay, B. C. Willard, J. J. Zayhowski, “Compact active hyperspectral imaging system for the detection of concealed targets,” in Detection and Remediation Technologies for Mines and Minelike Targets IV, A. C. Dubey, J. F. Harvey, J. T. Broach, R. E. Dugan, eds., Proc. SPIE3710, 144–153 (1999).
[CrossRef]

Kerekes, J. P.

B. Johnson, R. Joseph, M. L. Nischan, A. Newbury, J. P. Kerekes, H. T. Barclay, B. C. Willard, J. J. Zayhowski, “Compact active hyperspectral imaging system for the detection of concealed targets,” in Detection and Remediation Technologies for Mines and Minelike Targets IV, A. C. Dubey, J. F. Harvey, J. T. Broach, R. E. Dugan, eds., Proc. SPIE3710, 144–153 (1999).
[CrossRef]

Newbury, A.

B. Johnson, R. Joseph, M. L. Nischan, A. Newbury, J. P. Kerekes, H. T. Barclay, B. C. Willard, J. J. Zayhowski, “Compact active hyperspectral imaging system for the detection of concealed targets,” in Detection and Remediation Technologies for Mines and Minelike Targets IV, A. C. Dubey, J. F. Harvey, J. T. Broach, R. E. Dugan, eds., Proc. SPIE3710, 144–153 (1999).
[CrossRef]

Nischan, M. L.

B. Johnson, R. Joseph, M. L. Nischan, A. Newbury, J. P. Kerekes, H. T. Barclay, B. C. Willard, J. J. Zayhowski, “Compact active hyperspectral imaging system for the detection of concealed targets,” in Detection and Remediation Technologies for Mines and Minelike Targets IV, A. C. Dubey, J. F. Harvey, J. T. Broach, R. E. Dugan, eds., Proc. SPIE3710, 144–153 (1999).
[CrossRef]

Pagé, V.

F. Goudail, V. Pagé, Ph. Réfrégier, “Improving target detection with polarization diversity imaging,” in Second Symposium on Physics in Signal and Image Processing (SEE, Paris, 2001). pp. 340–345.

Pezzaniti, J. L.

J. L. Pezzaniti, R. A. Chipman, “Mueller matrix imaging polarimetry,” Opt. Eng. 34, 1558–1568 (1995).
[CrossRef]

Réfrégier, Ph.

C. Chesnaud, Ph. Réfrégier, V. Boulet, “Statistical region snake-based segmentation adapted to different physical noise models,” IEEE Trans. Pattern. Anal. Mach. Intell. 21, 1145–1157 (1999).
[CrossRef]

O. Germain, Ph. Réfrégier, “Snake-based method for the segmentation of objects in multichannel images degraded by speckle,” Opt. Lett. 24, 814–816 (1999).
[CrossRef]

Ph. Réfrégier, O. Germain, T. Gaidon, “Optimal snake segmentation of target and background with independent gamma density probabilities, application to speckled and preprocessed images,” Opt. Commun. 137, 382–388 (1997).
[CrossRef]

O. Germain, Ph. Réfrégier, “Optimal snake-based segmentation of a random luminance target on a spatially disjoint background,” Opt. Lett. 21, 1845–1847 (1996).
[CrossRef] [PubMed]

F. Goudail, V. Pagé, Ph. Réfrégier, “Improving target detection with polarization diversity imaging,” in Second Symposium on Physics in Signal and Image Processing (SEE, Paris, 2001). pp. 340–345.

Whitehead, V. S.

Willard, B. C.

B. Johnson, R. Joseph, M. L. Nischan, A. Newbury, J. P. Kerekes, H. T. Barclay, B. C. Willard, J. J. Zayhowski, “Compact active hyperspectral imaging system for the detection of concealed targets,” in Detection and Remediation Technologies for Mines and Minelike Targets IV, A. C. Dubey, J. F. Harvey, J. T. Broach, R. E. Dugan, eds., Proc. SPIE3710, 144–153 (1999).
[CrossRef]

Zayhowski, J. J.

B. Johnson, R. Joseph, M. L. Nischan, A. Newbury, J. P. Kerekes, H. T. Barclay, B. C. Willard, J. J. Zayhowski, “Compact active hyperspectral imaging system for the detection of concealed targets,” in Detection and Remediation Technologies for Mines and Minelike Targets IV, A. C. Dubey, J. F. Harvey, J. T. Broach, R. E. Dugan, eds., Proc. SPIE3710, 144–153 (1999).
[CrossRef]

Appl. Opt. (1)

IEEE Trans. Pattern. Anal. Mach. Intell. (1)

C. Chesnaud, Ph. Réfrégier, V. Boulet, “Statistical region snake-based segmentation adapted to different physical noise models,” IEEE Trans. Pattern. Anal. Mach. Intell. 21, 1145–1157 (1999).
[CrossRef]

Opt. Commun. (1)

Ph. Réfrégier, O. Germain, T. Gaidon, “Optimal snake segmentation of target and background with independent gamma density probabilities, application to speckled and preprocessed images,” Opt. Commun. 137, 382–388 (1997).
[CrossRef]

Opt. Eng. (1)

J. L. Pezzaniti, R. A. Chipman, “Mueller matrix imaging polarimetry,” Opt. Eng. 34, 1558–1568 (1995).
[CrossRef]

Opt. Lett. (2)

Other (7)

P. Clémenceau, S. Breugnot, L. Collot, “Polarization diversity imaging,” in Laser Radar Technology and Applications III, Proc. SPIE3380, 284–291 (1998).
[CrossRef]

J. W. Goodman, “;Laser Speckle and Related Phenomena,” in Statistical Properties of Laser Speckle Patterns, 9–75 [Springer-Verlag (Topics in Applied Physics Vol. 9), Heidelberg, 1975], pp. 9–75.

F. Goudail, V. Pagé, Ph. Réfrégier, “Improving target detection with polarization diversity imaging,” in Second Symposium on Physics in Signal and Image Processing (SEE, Paris, 2001). pp. 340–345.

R. A. Chipman, “Polarization diversity active imaging,” in Image Reconstruction and Restoration II, T. J. Schulz, ed., Proc. SPIE3170, 68–73 (1997).
[CrossRef]

B. Johnson, R. Joseph, M. L. Nischan, A. Newbury, J. P. Kerekes, H. T. Barclay, B. C. Willard, J. J. Zayhowski, “Compact active hyperspectral imaging system for the detection of concealed targets,” in Detection and Remediation Technologies for Mines and Minelike Targets IV, A. C. Dubey, J. F. Harvey, J. T. Broach, R. E. Dugan, eds., Proc. SPIE3710, 144–153 (1999).
[CrossRef]

S. Breugnot, Ph. Clémenceau, “Modeling and performances of a polarization active imager at lambda = 806 nm,” in Laser Radar Technology and Applications IV, G. W. Kamerman, C. Werner, eds., Proc. SPIE3707, 449–460 (1999).
[CrossRef]

J. W. Goodman, “The speckle effect in coherent imaging,” in Statistical Optics, (Wiley, New York, 1985), pp. 347–356.

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 (10)

Fig. 1
Fig. 1

Principle of active polarimetric imaging.

Fig. 2
Fig. 2

Example of real polarimetric images, with the TCI and the corresponding OSCI. (a) s 1(i, j), (b) s 2(i, j), (c) OSCI ρ(i, j).

Fig. 3
Fig. 3

Probability density function of the OSCI, when the channels s 1 and s2 are distributed with gamma law of order L = 50, and for three values of the parameter γ: from the left to the right, γ = 1, 3, 9.

Fig. 4
Fig. 4

Plot of the (γ1, γ2) plane.

Fig. 5
Fig. 5

(a) Binary mask w defining the object. (b) Channel s 1(i, j), γ1 = 1.2. The small white square is the initial shape of the snake. (c) Channel s 2(i, j), γ2 = 0.9. (d) OSCI computed from (b) and (c). (e) Result of the segmentation using the TCI snake on images (b) and (c). (f) Result of the segmentation using the OSCI snake on image (d).

Fig. 6
Fig. 6

Variation of the average number of misclassified pixels (ANMP) as a function of γ2 for γ1 = 1.4, by use of the TCI snake (◇) and the OSCI snake (+). The illumination is considered homogeneous, and the object is that represented in Fig. 5(a).

Fig. 7
Fig. 7

Estimation of the segmentation quality of the object in Fig. 5(a) under uniform and nonuniform illumination. For all configurations, γ1 = 1.2, γ2 = 0.9, and γ b = 3. (a) Average number of misclassified pixels for segmentation on TCI under uniform illumination (continuous horizontal line), under nonuniform illumination with different values of ℓ (◇) and for segmentation on OSCI (dotted horizontal line). (b) Median of the number of misclassified pixels for the same configurations.

Fig. 8
Fig. 8

CDF of the number of misclassified pixels (NMP) obtained after segmentation of the object in Figure 5(a) under uniform and nonuniform illumination. For all configurations, γ1 = 1.2, γ2 = 0.9, γ b = 3. (a) TCI under uniform illumination (◇); OSCI (+); Dotted curves (from top to bottom): TCI under nonuniform illumination for ℓ = 1; ℓ = 2; ℓ = 3; ℓ = 5; ℓ = 10. (b) TCI under uniform illumination (◇); TCI under nonuniform illumination for ℓ = 2 (+); for ℓ = 10 (□); for ℓ = 20 (×); for ℓ = 30 (△); for ℓ = 60 (*).

Fig. 9
Fig. 9

Examples of segmentation results on different realisations of scene for different configurations, with the corresponding value of the number of misclassified pixels (NMP). In all configurations, γ1 = 1.2, γ2 = 0.9, and in the case of nonuniform illumination, σ = 0.05. From the top: first row, TCI snake, uniform illumination; second row, TCI snake, nonuniform illumination with ℓ = 2; third row, TCI snake, ℓ = 10; fourth row, TCI snake, ℓ = 60; fifth row, OSCI snake, whatever the value of ℓ, γ b = 3.

Fig. 10
Fig. 10

Four examples of segmentation results on real polarimetric images. First column, s 1 with initial contour, second column, s 2; third column, s 1 with the result of TCI snake; fourth column, OSCI with the result of OSCI snake.

Equations (10)

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

ρi, j=s1i, j-s2i, js1i, j+s2i, j.
s1i, j=a1i, jwi, j+b2i, jw¯i, j,s2i, j=a2i, jwi, j+b2i, jw¯i, j,
we=arg maxw-Nawk=12log[mkaw-Nbwk=12log[mkbw,
mkaw=1Nawi,jw ski, jmkbw=1Nbw¯i,jw¯ ski, j.
we=arg maxw-Nawlogσkaw-Nbwlogσkbw,
σkαw2=1Nαwi,jwρi, j-1Nαwi,jw ρi, j2,
ρ=s1-s2s1+s2=μ1ar1-μ2ar2μ1ar1+μ2ar2=γar1-r2γar1+r2,
Cggm, n=σ2 exp-m2+n21/2.
s1i, j=θi, js1i, j,
s2i, j=θi, js2i, j,

Metrics