Abstract

An exclusively binarized method for the correlation calculation of image processing is presented with application to velocity measurements. The exclusively binarized correlation-calculation method is capable of determining the peak position of the correlation even with a small number of random-pattern images. Because the velocity is deduced from the movement of the correlation-peak position within a time interval, the method is suitable for the velocity measurement of random patterns. The method is applied to the measurement of the upper-atmospheric wind velocity by use of stellar scintillation patterns.

© 2001 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |
  • |

  1. J. Ohtsubo, “Velocity measurement using the time–space cross correlation of speckle patterns,” Opt. Commun. 34, 147–152 (1980).
    [CrossRef]
  2. F. Roddier, “The effect of atmospheric turbulence in optical astronomy,” in Progress in Optics XIX, E. Wolf, ed. (North-Holland, Amsterdam, 1981), pp. 281–376.
    [CrossRef]
  3. M. Azouit, J. Vernin, “Remote investigation of tropospheric turbulence by two-dimensional analysis of stellar scintillation,” J. Atmos. Sci. 37, 1550–1557 (1980).
    [CrossRef]
  4. J. L. Caccia, M. Azouit, J. Vernin, “Wind and CN2 profiling by single-star scintillation analysis,” Appl. Opt. 26, 1288–1294 (1987).
    [CrossRef] [PubMed]
  5. R. Li, M. Takabe, T. Aruga, “A new data acquisition system for measuring the movement of atmospheric speckle patterns,” Rev. Sci. Instrum. 70, 2136–2139 (1999).
    [CrossRef]
  6. S. Oya, T. Aruga, “Exclusively binarized correlation-calculation method for measuring random pattern movement,” Opt. Commun. 185, 239–248 (2000).
    [CrossRef]
  7. V. A. Klüeckers, N. J. Wooder, T. W. Nicholls, M. J. Adcock, I. Nunro, J. C. Dainty, “Profiling of atmospheric turbulence strength and velocity using a generalized SCIDAR,” Astron. Astrophys. Suppl. Ser. 130, 141–155 (1998).
    [CrossRef]

2000 (1)

S. Oya, T. Aruga, “Exclusively binarized correlation-calculation method for measuring random pattern movement,” Opt. Commun. 185, 239–248 (2000).
[CrossRef]

1999 (1)

R. Li, M. Takabe, T. Aruga, “A new data acquisition system for measuring the movement of atmospheric speckle patterns,” Rev. Sci. Instrum. 70, 2136–2139 (1999).
[CrossRef]

1998 (1)

V. A. Klüeckers, N. J. Wooder, T. W. Nicholls, M. J. Adcock, I. Nunro, J. C. Dainty, “Profiling of atmospheric turbulence strength and velocity using a generalized SCIDAR,” Astron. Astrophys. Suppl. Ser. 130, 141–155 (1998).
[CrossRef]

1987 (1)

1980 (2)

J. Ohtsubo, “Velocity measurement using the time–space cross correlation of speckle patterns,” Opt. Commun. 34, 147–152 (1980).
[CrossRef]

M. Azouit, J. Vernin, “Remote investigation of tropospheric turbulence by two-dimensional analysis of stellar scintillation,” J. Atmos. Sci. 37, 1550–1557 (1980).
[CrossRef]

Adcock, M. J.

V. A. Klüeckers, N. J. Wooder, T. W. Nicholls, M. J. Adcock, I. Nunro, J. C. Dainty, “Profiling of atmospheric turbulence strength and velocity using a generalized SCIDAR,” Astron. Astrophys. Suppl. Ser. 130, 141–155 (1998).
[CrossRef]

Aruga, T.

S. Oya, T. Aruga, “Exclusively binarized correlation-calculation method for measuring random pattern movement,” Opt. Commun. 185, 239–248 (2000).
[CrossRef]

R. Li, M. Takabe, T. Aruga, “A new data acquisition system for measuring the movement of atmospheric speckle patterns,” Rev. Sci. Instrum. 70, 2136–2139 (1999).
[CrossRef]

Azouit, M.

J. L. Caccia, M. Azouit, J. Vernin, “Wind and CN2 profiling by single-star scintillation analysis,” Appl. Opt. 26, 1288–1294 (1987).
[CrossRef] [PubMed]

M. Azouit, J. Vernin, “Remote investigation of tropospheric turbulence by two-dimensional analysis of stellar scintillation,” J. Atmos. Sci. 37, 1550–1557 (1980).
[CrossRef]

Caccia, J. L.

Dainty, J. C.

V. A. Klüeckers, N. J. Wooder, T. W. Nicholls, M. J. Adcock, I. Nunro, J. C. Dainty, “Profiling of atmospheric turbulence strength and velocity using a generalized SCIDAR,” Astron. Astrophys. Suppl. Ser. 130, 141–155 (1998).
[CrossRef]

Klüeckers, V. A.

V. A. Klüeckers, N. J. Wooder, T. W. Nicholls, M. J. Adcock, I. Nunro, J. C. Dainty, “Profiling of atmospheric turbulence strength and velocity using a generalized SCIDAR,” Astron. Astrophys. Suppl. Ser. 130, 141–155 (1998).
[CrossRef]

Li, R.

R. Li, M. Takabe, T. Aruga, “A new data acquisition system for measuring the movement of atmospheric speckle patterns,” Rev. Sci. Instrum. 70, 2136–2139 (1999).
[CrossRef]

Nicholls, T. W.

V. A. Klüeckers, N. J. Wooder, T. W. Nicholls, M. J. Adcock, I. Nunro, J. C. Dainty, “Profiling of atmospheric turbulence strength and velocity using a generalized SCIDAR,” Astron. Astrophys. Suppl. Ser. 130, 141–155 (1998).
[CrossRef]

Nunro, I.

V. A. Klüeckers, N. J. Wooder, T. W. Nicholls, M. J. Adcock, I. Nunro, J. C. Dainty, “Profiling of atmospheric turbulence strength and velocity using a generalized SCIDAR,” Astron. Astrophys. Suppl. Ser. 130, 141–155 (1998).
[CrossRef]

Ohtsubo, J.

J. Ohtsubo, “Velocity measurement using the time–space cross correlation of speckle patterns,” Opt. Commun. 34, 147–152 (1980).
[CrossRef]

Oya, S.

S. Oya, T. Aruga, “Exclusively binarized correlation-calculation method for measuring random pattern movement,” Opt. Commun. 185, 239–248 (2000).
[CrossRef]

Roddier, F.

F. Roddier, “The effect of atmospheric turbulence in optical astronomy,” in Progress in Optics XIX, E. Wolf, ed. (North-Holland, Amsterdam, 1981), pp. 281–376.
[CrossRef]

Takabe, M.

R. Li, M. Takabe, T. Aruga, “A new data acquisition system for measuring the movement of atmospheric speckle patterns,” Rev. Sci. Instrum. 70, 2136–2139 (1999).
[CrossRef]

Vernin, J.

J. L. Caccia, M. Azouit, J. Vernin, “Wind and CN2 profiling by single-star scintillation analysis,” Appl. Opt. 26, 1288–1294 (1987).
[CrossRef] [PubMed]

M. Azouit, J. Vernin, “Remote investigation of tropospheric turbulence by two-dimensional analysis of stellar scintillation,” J. Atmos. Sci. 37, 1550–1557 (1980).
[CrossRef]

Wooder, N. J.

V. A. Klüeckers, N. J. Wooder, T. W. Nicholls, M. J. Adcock, I. Nunro, J. C. Dainty, “Profiling of atmospheric turbulence strength and velocity using a generalized SCIDAR,” Astron. Astrophys. Suppl. Ser. 130, 141–155 (1998).
[CrossRef]

Appl. Opt. (1)

Astron. Astrophys. Suppl. Ser. (1)

V. A. Klüeckers, N. J. Wooder, T. W. Nicholls, M. J. Adcock, I. Nunro, J. C. Dainty, “Profiling of atmospheric turbulence strength and velocity using a generalized SCIDAR,” Astron. Astrophys. Suppl. Ser. 130, 141–155 (1998).
[CrossRef]

J. Atmos. Sci. (1)

M. Azouit, J. Vernin, “Remote investigation of tropospheric turbulence by two-dimensional analysis of stellar scintillation,” J. Atmos. Sci. 37, 1550–1557 (1980).
[CrossRef]

Opt. Commun. (2)

S. Oya, T. Aruga, “Exclusively binarized correlation-calculation method for measuring random pattern movement,” Opt. Commun. 185, 239–248 (2000).
[CrossRef]

J. Ohtsubo, “Velocity measurement using the time–space cross correlation of speckle patterns,” Opt. Commun. 34, 147–152 (1980).
[CrossRef]

Rev. Sci. Instrum. (1)

R. Li, M. Takabe, T. Aruga, “A new data acquisition system for measuring the movement of atmospheric speckle patterns,” Rev. Sci. Instrum. 70, 2136–2139 (1999).
[CrossRef]

Other (1)

F. Roddier, “The effect of atmospheric turbulence in optical astronomy,” in Progress in Optics XIX, E. Wolf, ed. (North-Holland, Amsterdam, 1981), pp. 281–376.
[CrossRef]

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 diagram of the scintillation-pattern measurement system.

Fig. 2
Fig. 2

Concept of the relation between an ICCD frame’s data pair and a synchronously obtained PMT record. One thousand such data pairs are acquired for ensemble averaging in the correlation calculation.

Fig. 3
Fig. 3

Histogram of the observed scintillation intensity of the ICCD pixel that corresponds to the PMT’s pinhole position. The abscissa is in ADUs (8 bits). The width of one bin is 4 ADUs.

Fig. 4
Fig. 4

Relation between the intensities of the ICCD pixel that corresponds to the PMT’s pinhole position and the PMT outputs at a delay time of τ = 0 ms. Both axes are in ADUs (8 bits for the ICCD and 12 bits for the PMT). The solid curve represents the results of the linear least-squares fitting. The dotted curves indicate the tolerable-error range that was used in the binarizing method.

Fig. 5
Fig. 5

Two-dimensional correlation strength between the intensities of the ICCD pixels and the PMT output at a delay time of τ = 0 ms. Thus the peak (indicated by the circle) appears at the position that corresponds to the PMT’s pinhole: the peak position as calculated by use of (a) the general method and (b) the EBM. All 854 data pairs that passed the saturation check were used. In both (a) and (b) the size of the image was 64 pixels × 64 pixels with a pixel scale of 2.8 mm/pixel. The pixel value is normalized so that the background level is zero (black) and the maximum value is unity (white). Marginal pixels are drawn in black because the number of adjacent pixels was insufficient for the smoothing window to have an effect during preprocessing. Some pixels at the lower edges and the corners are also set to black to mask the image of the secondary mirror and the spider of the telescope.

Fig. 6
Fig. 6

Same as in Fig. 5 but with a delay time of τ = 1 ms: The peak position was calculated by use of (a) the general method and (b) the EBM. The circle is shifted to the peak position at the delay time. The peak appears at the position from which the patterns of the ICCD pixels moved to the pinhole position at the delay time of 1 ms.

Fig. 7
Fig. 7

Same as in Fig. 5 (τ = 0 ms), but only the first 25 data pairs were used for the correlation calculation. The peak position was calculated by use of (a) the general method and (b) the EBM.

Fig. 8
Fig. 8

Same as in Fig. 6 (τ = 1 ms), but only the first 25 data pairs were used for the correlation calculation. The peak position was calculated by use of (a) the general method and (b) the EBM.

Fig. 9
Fig. 9

Deviation of the peak position from the correct position for each subset composed of 25 data pairs. The solid curve represents the results of the EBM, and the dashed curve the general method. The delay time is τ = 1 ms.

Fig. 10
Fig. 10

Schematically drawn example of a scintillation-pattern image that causes false peaks in the two-dimensional correlation strength. The pattern moves with the same velocity v as does the projected wind velocity, and only the smaller peak is observed by the PMT at the delay time τ.

Fig. 11
Fig. 11

Noise-reduction effect of the EBM: The two-dimensional correlation strengths of a subset that generates the largest deviation in Fig. 9 (the 13th subset) is shown. The delay time is τ = 1 ms. (a), (b), and (c) were calculated with the general method; (d) and (e) were calculated with the EBM. All images are normalized in the same manner as was done for Fig. 5. Each circle indicates the correct peak position, as in Fig. 6, and that position is plotted in either black or white for readability against the background color. Each cross indicates a peak position that was falsely determined by the general method for this data subset, also plotted in black or white, as for the correct peak position. (a) The peak position obtained by use of all 25 data pairs but showing a spurious peak to be higher than the peak near the correct position. (b) The harmful image that generates the spurious peak shown in (a) of the stellar scintillation pattern. (c) The result obtained by use of the general method and only 24 data pairs, i.e., by exclusion of the harmful image. (d) Only the correct peak appears with the EBM even though all 25 data pairs were used. (e) The binarized (by the EBM) image of the harmful image shown in (b). By virtue of the EBM the large peak in the harmful image was successfully excluded, and the correct peak remains.

Equations (6)

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

Cˆx; τ=IICCDx×IPMTτIICCDxIPMTτ
Cx; τ=IICCDx-IICCDx×IPMTτ-IPMTτ.
Cˆx; τ-1=IICCDx-IICCDxIPMTτ-IPMTτIICCDxIPMTτ=Cx; τmI2.
SNRG=11+21+σEσI221/21-1N1/2,
Bix; τ=1IPMTiτ-δIICCDixIPMTiτ+δ0otherwise.
SNRB=PS-PBPS1-PS+PB1-PB1/2.

Metrics