Abstract

To accomplish color constancy the illuminant color needs to be discounted from the light reflected from surfaces. Some strategies for discounting the illuminant color use statistics of luminance and chromaticity distribution in natural scenes. In this study we showed whether color constancy exploits the potential cue that was provided by the luminance balance of differently colored surfaces. In our experiments we used six colors: bright and dim red, green, and blue, as surrounding stimulus colors. In most cases, bright colors were set to be optimal colors. They were arranged among 60 hexagonal elements in close-packed structure. The center element served as the test stimulus. The observer adjusted the chromaticity of the test stimulus to obtain a perceptually achromatic surface. We used simulated black body radiations of 3000 (or 4000), 6500, and 20000 K as test illuminants. The results showed that the luminance balance of surfaces with no chromaticity shift had clear effects on the observer’s achromatic setting, which was consistent with our hypothesis on estimating the scene illuminant based on optimal colors.

© 2012 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |
  • |

  1. D. H. Foster, “Color constancy,” Vis. Res. 51, 674–700 (2011).
    [CrossRef]
  2. G. Buchsbaum, “A spatial processor model for object colour perception,” J. Franklin Inst. 310, 1–26 (1980).
    [CrossRef]
  3. E. H. Land, “Recent advances in Retinex theory,” Vis. Res. 26, 7–21 (1986).
    [CrossRef]
  4. R. Brown, “Background and illuminants: The yin and yang of colour constancy,” in Colour Perception: Mind and the Physical World, R. Mausfeld and D. Heyer, eds. (Oxford University, 2003).
  5. J. Golz and D. I. A. MacLeod, “Influence of scene statistics on colour constancy,” Nature 415, 637–640 (2002).
    [CrossRef]
  6. D. L. Ruderman, T. W. Cronin, and C. Chiao, “Statistics of cone responses to natural images: implication for visual coding,” J. Opt. Soc. Am. A 15, 2036–2045 (1998).
    [CrossRef]
  7. L. T. Maloney and B. A. Wandell, “Color constancy: a method for recovering surface spectral reflectance,” J. Opt. Soc. Am. A 3, 29–33 (1986).
    [CrossRef]
  8. D. H. Brainard and W. T. Freeman, “Bayesian color constancy,” J. Opt. Soc. Am. A 14, 1393–1411 (1997).
    [CrossRef]
  9. D. A. Forsyth, “A novel algorithm for color constancy,” Int. J. Comput. Vis. 5, 5–36 (1990).
    [CrossRef]
  10. G. D. Finalayson, P. M. Hubel, and S. Hordley, “Color by correlation,” in Proceedings of the Fifth Color Imaging Conference (Society for Imaging Science and Technology, 1997), pp. 6–11.
  11. G. Wyszecki and W. S. Stiles, Color Science, 2nd ed. (Wiley, 1982).
  12. J. J. Koenderink and A. J. van Doorn, “Perspectives on colour space,” in Colour Perception: Mind and the Physical World, D. Mausfeld and D. Heyer, eds. (Oxford University, 2003).
  13. D. B. Judd, “Hue saturation and lightness of surface colors with chromatic illumination,” J. Opt. Soc. Am. 30, 2–32 (1940).
    [CrossRef]
  14. A. Gilchrist, Seeing Black and White (Oxford University, 2006).
  15. E. H. Land and J. J. McCann, “Lightness and Retinex theory,” J. Opt. Soc. Am. 61, 1–11 (1971).
    [CrossRef]
  16. A. Stockman, D. I. A. MacLeod, and N. E. Johnson, “Spectral sensitivities of the human cones,” J. Opt. Soc. Am. A 10, 2491–2621 (1993).
    [CrossRef]
  17. D. I. A. MacLeod and R. M. Boynton, “Chromaticity diagram showing cone excitation by stimuli of equal luminance,” J. Opt. Soc. Am. 69, 1183–1186 (1979).
    [CrossRef]
  18. K. Uchikawa, K. Koida, T. Meguro, Y. Yamauchi, and I. Kuriki, “Brightness, not luminance, determines transition from the surface-color to the aperture-color mode for colored lights,” J. Opt. Soc. Am. A 18, 737–746 (2001).
    [CrossRef]
  19. J. M. Speigle and D. H. Brainard, “Luminosity thresholds: effects of test chromaticity and ambient illumination,” J. Opt. Soc. Am. A 13, 436–451 (1996).
    [CrossRef]
  20. S. Tominaga, S. Ebisui, and B. A. Wandell, “Scene illuminant classification: brighter is better,” J. Opt. Soc. Am. A 18, 55–64 (2001).
    [CrossRef]
  21. To further examine this point, it may be useful to note the luminance balances actually used in experiment 4. Under the 4000 K illuminant, surface chromaticities were shifted redward, so that to achieve an L, M, S average equal to the equal energy white, the luminance balance had to be adjusted by dimming the R colors, yielding an R∶B luminance ratio of 1∶3.7; this is much lower than the ratio (1∶0.69) characteristic of optimal colors of those chromaticities under 4000 K (and slightly lower even than the ratio (1∶2.39) that we obtain for optimal colors of those chromaticities under 20000 K). Similarly, the R∶B luminance ratio under our 20000 K illuminant condition (R∶B=1∶0.149) was far higher than would be appropriate for optimal colors of that chromaticity under 20000 K (R∶B=1∶1.123); it was close to, but still more extreme than, that of such optimal colors observed under 4000 K (R∶B=1∶0.339).

2011 (1)

D. H. Foster, “Color constancy,” Vis. Res. 51, 674–700 (2011).
[CrossRef]

2002 (1)

J. Golz and D. I. A. MacLeod, “Influence of scene statistics on colour constancy,” Nature 415, 637–640 (2002).
[CrossRef]

2001 (2)

1998 (1)

1997 (1)

1996 (1)

1993 (1)

1990 (1)

D. A. Forsyth, “A novel algorithm for color constancy,” Int. J. Comput. Vis. 5, 5–36 (1990).
[CrossRef]

1986 (2)

1980 (1)

G. Buchsbaum, “A spatial processor model for object colour perception,” J. Franklin Inst. 310, 1–26 (1980).
[CrossRef]

1979 (1)

1971 (1)

1940 (1)

Boynton, R. M.

Brainard, D. H.

Brown, R.

R. Brown, “Background and illuminants: The yin and yang of colour constancy,” in Colour Perception: Mind and the Physical World, R. Mausfeld and D. Heyer, eds. (Oxford University, 2003).

Buchsbaum, G.

G. Buchsbaum, “A spatial processor model for object colour perception,” J. Franklin Inst. 310, 1–26 (1980).
[CrossRef]

Chiao, C.

Cronin, T. W.

Ebisui, S.

Finalayson, G. D.

G. D. Finalayson, P. M. Hubel, and S. Hordley, “Color by correlation,” in Proceedings of the Fifth Color Imaging Conference (Society for Imaging Science and Technology, 1997), pp. 6–11.

Forsyth, D. A.

D. A. Forsyth, “A novel algorithm for color constancy,” Int. J. Comput. Vis. 5, 5–36 (1990).
[CrossRef]

Foster, D. H.

D. H. Foster, “Color constancy,” Vis. Res. 51, 674–700 (2011).
[CrossRef]

Freeman, W. T.

Gilchrist, A.

A. Gilchrist, Seeing Black and White (Oxford University, 2006).

Golz, J.

J. Golz and D. I. A. MacLeod, “Influence of scene statistics on colour constancy,” Nature 415, 637–640 (2002).
[CrossRef]

Hordley, S.

G. D. Finalayson, P. M. Hubel, and S. Hordley, “Color by correlation,” in Proceedings of the Fifth Color Imaging Conference (Society for Imaging Science and Technology, 1997), pp. 6–11.

Hubel, P. M.

G. D. Finalayson, P. M. Hubel, and S. Hordley, “Color by correlation,” in Proceedings of the Fifth Color Imaging Conference (Society for Imaging Science and Technology, 1997), pp. 6–11.

Johnson, N. E.

Judd, D. B.

Koenderink, J. J.

J. J. Koenderink and A. J. van Doorn, “Perspectives on colour space,” in Colour Perception: Mind and the Physical World, D. Mausfeld and D. Heyer, eds. (Oxford University, 2003).

Koida, K.

Kuriki, I.

Land, E. H.

E. H. Land, “Recent advances in Retinex theory,” Vis. Res. 26, 7–21 (1986).
[CrossRef]

E. H. Land and J. J. McCann, “Lightness and Retinex theory,” J. Opt. Soc. Am. 61, 1–11 (1971).
[CrossRef]

MacLeod, D. I. A.

Maloney, L. T.

McCann, J. J.

Meguro, T.

Ruderman, D. L.

Speigle, J. M.

Stiles, W. S.

G. Wyszecki and W. S. Stiles, Color Science, 2nd ed. (Wiley, 1982).

Stockman, A.

Tominaga, S.

Uchikawa, K.

van Doorn, A. J.

J. J. Koenderink and A. J. van Doorn, “Perspectives on colour space,” in Colour Perception: Mind and the Physical World, D. Mausfeld and D. Heyer, eds. (Oxford University, 2003).

Wandell, B. A.

Wyszecki, G.

G. Wyszecki and W. S. Stiles, Color Science, 2nd ed. (Wiley, 1982).

Yamauchi, Y.

Int. J. Comput. Vis. (1)

D. A. Forsyth, “A novel algorithm for color constancy,” Int. J. Comput. Vis. 5, 5–36 (1990).
[CrossRef]

J. Franklin Inst. (1)

G. Buchsbaum, “A spatial processor model for object colour perception,” J. Franklin Inst. 310, 1–26 (1980).
[CrossRef]

J. Opt. Soc. Am. (3)

J. Opt. Soc. Am. A (7)

Nature (1)

J. Golz and D. I. A. MacLeod, “Influence of scene statistics on colour constancy,” Nature 415, 637–640 (2002).
[CrossRef]

Vis. Res. (2)

D. H. Foster, “Color constancy,” Vis. Res. 51, 674–700 (2011).
[CrossRef]

E. H. Land, “Recent advances in Retinex theory,” Vis. Res. 26, 7–21 (1986).
[CrossRef]

Other (6)

R. Brown, “Background and illuminants: The yin and yang of colour constancy,” in Colour Perception: Mind and the Physical World, R. Mausfeld and D. Heyer, eds. (Oxford University, 2003).

G. D. Finalayson, P. M. Hubel, and S. Hordley, “Color by correlation,” in Proceedings of the Fifth Color Imaging Conference (Society for Imaging Science and Technology, 1997), pp. 6–11.

G. Wyszecki and W. S. Stiles, Color Science, 2nd ed. (Wiley, 1982).

J. J. Koenderink and A. J. van Doorn, “Perspectives on colour space,” in Colour Perception: Mind and the Physical World, D. Mausfeld and D. Heyer, eds. (Oxford University, 2003).

A. Gilchrist, Seeing Black and White (Oxford University, 2006).

To further examine this point, it may be useful to note the luminance balances actually used in experiment 4. Under the 4000 K illuminant, surface chromaticities were shifted redward, so that to achieve an L, M, S average equal to the equal energy white, the luminance balance had to be adjusted by dimming the R colors, yielding an R∶B luminance ratio of 1∶3.7; this is much lower than the ratio (1∶0.69) characteristic of optimal colors of those chromaticities under 4000 K (and slightly lower even than the ratio (1∶2.39) that we obtain for optimal colors of those chromaticities under 20000 K). Similarly, the R∶B luminance ratio under our 20000 K illuminant condition (R∶B=1∶0.149) was far higher than would be appropriate for optimal colors of that chromaticity under 20000 K (R∶B=1∶1.123); it was close to, but still more extreme than, that of such optimal colors observed under 4000 K (R∶B=1∶0.339).

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 (14)

Fig. 1.
Fig. 1.

The chromaticity and luminance distribution of all optimal colors under 20000 K, 6500 K, and 3000 K illuminant. The abscissa represents redness in the MacLeod–Boynton chromaticity diagram.

Fig. 2.
Fig. 2.

The chromaticity and luminance distribution of optimal colors and 574 natural objects measured by Brown under (a): 20000 K , (b): 6500 K, and (c): 3000 K illuminant. The abscissa represents redness in the MacLeod–Boynton chromaticity diagram.

Fig. 3.
Fig. 3.

An example of the stimulus spatial configuration used in the experiments. The surrounding field consisted of 60 hexagons of bright and dim R, G, B colors. The center hexagon was used as the test field.

Fig. 4.
Fig. 4.

Chromaticities of test illuminants, mean chromaticities of surrounding R, G, B colors and means of L, M, S cone responses of surrounding R, G, B colors used in experiment 1. Stimulus condition: 1 (20000 K), 5 (6500 K) and 9 (3000 K).

Fig. 5.
Fig. 5.

Observer’s achromatic settings obtained in experiment 1 in three test luminance conditions (L=0.1, 0.25, and 0.5) for observer (a) KU and (b) YK. Closed symbols represent means of settings and small dots show settings for each trial. Stimulus conditions: 1 (diamond), 5 (circle), and 9 (square). Positions of illuminant: 20000 K (open diamond), 6500 K (open circle), and 3000 K (open square). Stimulus condition: 1 (20000 K), 5 (6500 K) and 9 (3000 K).

Fig. 6.
Fig. 6.

Constancy indexes (CIs) for two observers obtained in experiment 1. Conditions: 1-5-9 (filled circle), 4-5-6 (filled triangle) and 2-5-8 (open triangle).

Fig. 7.
Fig. 7.

Chromaticities of test illuminants, mean chromaticities of surrounding R, G, B colors and means of L, M, S cone responses of surrounding R, G, B colors used in experiment 2. Stimulus condition: 1 (20000 K), 5 (6500 K), and 9 (3000 K).

Fig. 8.
Fig. 8.

CIs for two observers obtained in experiment 2. Conditions: 1-5-9 (filled circle), 4-5-6 (filled triangle), and 2-5-8 (open triangle).

Fig. 9.
Fig. 9.

Chromaticities of test illuminants, mean chromaticities of surrounding R, G, B colors and means of L, M, S cone responses of surrounding R, G, B colors used in experiment 3. Stimulus condition: 1 (20000 K), 5 (6500 K), and 9 (3000 K).

Fig. 10.
Fig. 10.

CIs for two observers obtained in experiment 3. Conditions: 1-5-9 (filled circle), 4-5-6 (filled triangle), and 2-5-8 (open triangle).

Fig. 11.
Fig. 11.

Ratio of CI for luminance balance and chromaticity shift obtained in experiment 1, 2, 3 in the condition of 1-5-9.

Fig. 12.
Fig. 12.

Observer’s achromatic settings obtained in experiment 4 for four observers. The same test illuminant was used both for bright and dim R, G, B colors. Test luminance was 0.25.

Fig. 13.
Fig. 13.

CIs for four observers obtained in experiment 4. Illuminants: 20000 K and 4000 K.

Fig. 14.
Fig. 14.

Standard deviations (SDs) of observer’s settings in all experiments. (a) Observer KU. (b) Observer YK. The SDs are separately shown in redness and blueness directions, for each test luminance, averaged across all stimulus conditions.

Tables (5)

Tables Icon

Table 1. Combination of Test Illuminants for Separately Illuminated Surrounding Bright and Dim R, G, B Colors, with Numbers Representing the Conditions of Illuminants for Bright and Dim R, G, B Colors

Tables Icon

Table 2. MacLeod–Boynton Chromaticity Coordinates and Luminance of R, G, B Colors Used in Experiment 1 (Luminance: 0.5=28.6cd/m2)

Tables Icon

Table 3. MacLeod–Boynton Chromaticity Coordinates and Luminance of R, G, B Colors Used in Experiment 2 (Luminance: 0.5=28.6cd/m2)

Tables Icon

Table 4. MacLeod–Boynton Chromaticity Coordinates and Luminance of R, G, B Colors Used in Experiment 3 (Luminance: 0.5=28.6cd/m2)

Tables Icon

Table 5. MacLeod–Boynton Chromaticity Coordinates and Luminance of R, G, B Colors Used in Experiment 4 (Luminance: 0.5=28.6cd/m2)

Equations (1)

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

CI=ds(ST-6500K)/di(PT-6500K).

Metrics