OSA's Digital Library

Journal of the Optical Society of America A

Journal of the Optical Society of America A

| OPTICS, IMAGE SCIENCE, AND VISION

  • Vol. 16, Iss. 3 — Mar. 1, 1999
  • pp: 728–741

Local versus global contrasts in texture segregation

Andrei Gorea and Thomas V. Papathomas  »View Author Affiliations


JOSA A, Vol. 16, Issue 3, pp. 728-741 (1999)
http://dx.doi.org/10.1364/JOSAA.16.000728


View Full Text Article

Acrobat PDF (609 KB)





Browse Journals / Lookup Meetings

Browse by Journal and Year


   


Lookup Conference Papers

Close Browse Journals / Lookup Meetings

Article Tools

Share
Citations

Abstract

In a texture pair (TP) yielding a vertical or horizontal edge, the local (luminance or color) contrast or the local orientation of the individual textels is traded off with the global strength of the luminance-, color-, or orientation-defined TP edge so as to keep the latter at the detection threshold. Local and global contrasts are defined along the same (within-domain conditions) or along distinct physical dimensions (transdomain conditions). In the latter case local luminance or color contrast is traded off against global orientation. In all cases TP’s are presented for 66.7 or 333.3 ms. Textels differ from the background in either luminance or color so that the TP’s are respectively equichromatic or equiluminant. TP edge strength is modulated by means of swapping variable proportions of textels between the two textures in the TP. The observed local–global relationships are fitted with a version of the equivalent noise model for contrast coding modified to include the presentation time factor. The extension of the standard model in the time domain is meant to allow comparison between equivalent noise estimates for variable duration stimuli. Model fits of the within-domain data yield equivalent noise energy values significantly different for color- and luminance-defined TP’s but are not applicable for the transdomain experiments, which indicates that global orientation processing is independent of both local luminance and local color contrast insofar as the latter are above the detection threshold. Finally, this study points to the equivalence among the local–global, the equivalent noise, and the statistical approaches to texture segregation.

© 1999 Optical Society of America

OCIS Codes
(330.4060) Vision, color, and visual optics : Vision modeling
(330.5510) Vision, color, and visual optics : Psychophysics

Citation
Andrei Gorea and Thomas V. Papathomas, "Local versus global contrasts in texture segregation," J. Opt. Soc. Am. A 16, 728-741 (1999)
http://www.opticsinfobase.org/josaa/abstract.cfm?URI=josaa-16-3-728


Sort:  Author  |  Year  |  Journal  |  Reset

References

  1. A. Gorea, “Visual texture,” in Early Vision and Beyond, T. V. Papathomas, C. Chubb, A. Gorea, and E. Kowler, eds. (MIT, Cambridge, Mass., 1995), pp. 55–57.
  2. J. R. Bergen, “Theories of visual texture perception,” in Spatial Vision, D. Regan, ed. (CRC Press, New York, 1991), pp. 114–139.
  3. J. R. Bergen and M. S. Landy, “Computational modeling of visual texture segregation,” in Computational Models of Visual Processing, M. S. Landy and J. A. Movshon, eds. (MIT, Cambridge, Mass., 1991), pp. 253–271.
  4. D. Sagi, “The psychophysics of texture segmentation,” in Early Vision and Beyond, T. V. Papathomas, C. Chubb, A. Gorea, and E. Kowler, eds. (MIT, Cambridge, Mass., 1995), pp. 70–78.
  5. D. Navon, “Forest before trees: the precedence of global features in visual perception,” Cogn. Psychol. 9, 353–383 (1977).
  6. D. Navon, “The forest revisited: more on global precedence,” Psychol. Res. 43, 1–32 (1981).
  7. R. A. Kinchla and J. M. Wolfe, “The order of visual processing: ‘top-down, ’ ‘bottom-up’ or ‘middle-out, ’ ” Percept. Psychophys. 25, 225–231 (1979).
  8. D. Marr, Vision (Freeman, San Francisco, Calif., 1982).
  9. B. Julesz, “Visual pattern discrimination,” IRE Trans. Inf. Theory 8, 84–92 (1962).
  10. J. Beck, “Perceptual grouping produced by changes in orientation and shape,” Science 154, 538–540 (1966).
  11. B. Julesz, “Textons, the elements of texture perception and their interactions,” Nature (London) 290, 91–97 (1981).
  12. R. Rosenholtz, “Texture and image segmentation,” Perception Suppl. 26, 111 (1997).
  13. C. Chubb and M. S. Landy, “Orthogonal distribution analysis: a new approach to the study of texture perception,” in Computational Models of Visual Processing, M. S. Landy and J. A. Movshon, eds. (MIT, Cambridge, Mass., 1990), pp. 291–301.
  14. S. C. Dakin and R. J. Watt, “The computation of orientation statistics from visual texture,” Vision Res. 37, 3181–3192 (1997).
  15. D. G. Pelli, “Effects of visual noise,” Ph.D. dissertation (University of Cambridge, Cambridge, UK, 1981).
  16. D. G. Pelli, “The quantum efficiency of vision,” in Vision: Coding and Efficiency, C. Blakemore, ed. (Cambridge U. Press, New York, 1990), pp. 3–24.
  17. A. J. Ahumada, Jr. and A. B. Watson, “Equivalent-noise model for contrast detection and discrimination,” J. Opt. Soc. Am. A 2, 1133–1139 (1985).
  18. A. J. Ahumada, Jr., “Putting the visual system noise back in the picture,” J. Opt. Soc. Am. A 4, 2372–2378 (1987).
  19. A. J. Ahumada, Jr. and B. L. Beard, “Parafoveal target detectability reversal predicted by local luminance and gain control,” Invest. Ophthalmol. Visual Sci. Suppl. 38, S380 (1997).
  20. H. B. Barlow, “Retinal noise and absolute threshold,” J. Opt. Soc. Am. 46, 634–639 (1956).
  21. H. B. Barlow, “Increment thresholds at low intensities considered as signal/noise discrimination,” J. Physiol. (London) 136, 469–488 (1957).
  22. This approach appears to characterize the rather complex visual processing stream without loss of generality. Threshold predictions based on this single, generic front-end filter (whose Fourier transform is the contrast transfer function of the visual system as a whole) are quite comparable with those yielded by more realistic models involving the parallel processing of the input image through a bank of front-end filters tuned to different spatial frequencies and orientations.1718
  23. Assume that there are 2 textels of luminance L0+ΔL and L0−ΔL, with L0 being the luminance of the background. Their contrast relative both to L0 and to each other is c=ΔL/L0. The overall luminance in a TP is LT1=pd(L0+ΔL)+(1−p)d(L0−ΔL)+(1−d)L0= dΔL(2p−1)+L0 for one texture and LT2=pd(L0− ΔL)+(1−p)d(L0+ΔL)+(1−d)L0=−dΔL(2p−1)+ L0 for the other texture. The global TP contrast is given by (LT1−LT2)/(LT1+LT2)=c(P−1)d, with P= 2p, so that 50% swapped textels (p=0.5) yields 0% global contrast. As noted in the text, the status of the density parameter, d, is not clear when one is averaging second-order features such as orientation; mean orientation contrast could well be a nonmonotonic function of density.
  24. In experiments in which presentation time, T, is a parameter (as was the case here), local contrast should be replaced by local energy, which can be approximated by CLT insofar as T is smaller than the temporal integration constant (see Section 6).
  25. B. A. Dosher and Z.-L. Lu, “Mechanisms of perceptual learning,” Invest. Ophthalmol. Visual Sci. Suppl. 39, S912 (1998).
  26. Z.-L. Lu and B. A. Dosher, “External noise distinguishes attention mechanisms,” Vision Res. 38, 1183–1198 (1998).
  27. In the terminology of Gorea and Papathomas, 2829 these TP’s are of the type luminance (or color) across orientation (L×O or C×O) and orientation across luminance (or color) (O×L or O×C). More generally, A×B TP’s are those whose edges are defined by two (or more) values of attribute A with two (or more) values of the remaining attribute, B, randomized over the whole TP.
  28. A. Gorea and T. V. Papathomas, “Texture segregation by chromatic and achromatic visual pathways: an analogy with motion perception,” J. Opt. Soc. Am. 8, 386–393 (1991).
  29. A. Gorea and T. V. Papathomas, “Extending a class of motion stimuli to study multiattribute texture perception,” Behav. Res. Methods Instrum. 23, 5–8 (1991).
  30. Once the extraction of the TP edge is achieved, specifying edge orientation (vertical or horizontal) or edge location (±3.2-deg location discrimination) is a trivial task in the sense that these tasks do not limit texture segregation: They both yield 100% performance, provided that the edge itself is 100% visible; thus performances below 100% in any of the two tasks reflect only edge extraction limitations. Moreover, in the remainder of this paper the only comparative discussion between the luminance/color and the orientation data shall bear on the generality of the generic equivalent noise model and not on the details of the fits. The equivalent noises estimated for the luminance/color domains, on the one hand, and for the orientation domain, on the other, are not commensurable and therefore are not to be related even when estimated by means of identical procedures.
  31. A. Gorea, T. V. Papathomas, and I. Kovacs, “Motion perception with spatiotemporally matched chromatic and achromatic information reveals a ‘slow’ and a ‘fast’ motion system,” Vision Res. 33, 2515–2534 (1993).
  32. T. V. Papathomas, A. Gorea, A. Feher, and T. E. Conway, “Attention-based texture segregation,” Percept. Psychophys. (to be published).
  33. C. Agonie and A. Gorea, “Equivalent luminance contrast of red–green drifting stimuli: dependency on luminance–color interactions and on the psychophysical task,” J. Opt. Soc. Am. A 10, 1341–1352 (1993).
  34. The distribution of the two-valued attributes (luminance, color, or orientation) in the textures used here is binomial. The upper confidence limit of the binomial distribution with respect to p is given by p+zpq/N, where q=1−p; N is the number of events considered (in this particular case, the number of textels); and z, the standard score in the normal distribution, with the upper α/2 proportion of cases cut off. For p=0.5, N=400, and α=0.05, this expression yields a p value of 0.549. When converted into P−1 = 2p−1 (see Ref. 23), it becomes 0.098. The linear function delimiting the shaded area is given by 0.098CLd, which is the limit below which the measured global contrast at threshold cannot be distinguished from random variations of the binomial distribution around p=0.5.
  35. A. Gorea and C. W. Tyler, “New look at Bloch’s law for contrast,” J. Opt. Soc. Am. A 3, 52–61 (1986).
  36. D. H. Kelly and D. van Norren, “Two-band model of heterochromatic flicker,” J. Opt. Soc. Am. 67, 1081–1091 (1977).
  37. S. S. Wolfson and M. S. Landy, “Examining edge- and region-based texture analysis mechanisms,” Vision Res. 38, 439–446 (1998).
  38. E. Matin and A. Drivas, “Acuity for orientation measured with a sequential recognition task and signal detection methods,” Percept. Psychophys. 25, 161–168 (1979).
  39. D. W. Heeley and B. Timney, “Meridional anisotropies of orientation discrimination for sine wave gratings,” Vision Res. 28, 337–344 (1988).
  40. B. G. Smith and J. P. Thomas, “Why are some spatial discriminations independent of contrast,” J. Opt. Soc. Am. A 6, 713–724 (1989).
  41. D. W. Heeley and H. M. Buchanan-Smith, “Recognition of stimulus orientation,” Vision Res. 30, 1429–1437 (1990).
  42. R. E. Näsänen, H. T. Kukkonen, and J. M. Rovamo, “Modeling spatial integration and contrast invariance in visual pattern discrimination,” Invest. Ophthalmol. Visual Sci. 37, 260–266 (1997).
  43. Obviously, the process responsible for the decision as to whether a signal is present receives inputs from both the external signal and noise and the internal noise.16 Simulating it with an integrator that acts on a window of fixed duration is only a first approximation toward incorporating the effect of time into the model. It is assumed that the decision stage acts on the probabilistic sum of the outputs of several temporal integrators, each characterized by a critical integration time, τ.35 For obvious reasons, TD must be larger than the average integration time τ.
  44. D. Regan, “Orientation discrimination for bars defined by orientation texture,” Perception 24, 1131–1138 (1995).
  45. H. C. Nothdurft, “Sensitivity for structure gradient in texture discrimination tasks,” Vision Res. 25, 1957–1968 (1985).
  46. D. Sagi and B. Julesz, “Short-range limitation on detection of feature differences,” Spatial Vision 2, 39–49 (1987).
  47. P. Hammond, D. P. Andrews, and C. R. James, “Invariance of orientational and directional tuning in visual cortical cells of the adult cat,” Brain Res. 96, 56–59 (1975).
  48. G. Sclar and R. D. Freeman, “Orientation selectivity in the cat’s striate cortex is invariant with stimulus contrast,” Exp. Brain Res. 46, 457–461 (1982).
  49. L. Chao-yi and O. Creutzfeldt, “The representation of contrast and other stimulus parameters by single neurons in area 17 of the cat,” Pfluegers Arch. 401, 304–314 (1984).
  50. J. M. Foley, “Human luminance pattern-vision mechanisms: masking experiments require a new model,” J. Opt. Soc. Am. A 11, 1710–1719 (1994).
  51. J. M. Foley and C.-C. Chen, “Analysis of the effect of pattern adaptation on pattern pedestal effects: a two-process model,” Vision Res. 37, 2779–2788 (1997).
  52. J. M. Foley and W. Schwartz, “Spatial attention: effect of position uncertainty and number of distractor patterns on the threshold-versus-contrast function for contrast discrimination,” J. Opt. Soc. Am. A 15, 1036–1047 (1998).
  53. H. R. Wilson, “A transducer function for threshold and suprathreshold human vision,” Biol. Cybern. 38, 171–178 (1980).
  54. G. Sclar, J. H. R. Maunsell, and P. Lennie, “Coding of image contrast in central visual pathways of the macaque monkey,” Vision Res. 30, 1–10 (1990).

Cited By

Alert me when this paper is cited

OSA is able to provide readers links to articles that cite this paper by participating in CrossRef's Cited-By Linking service. CrossRef includes content from more than 3000 publishers and societies. In addition to listing OSA journal articles that cite this paper, citing articles from other participating publishers will also be listed.


« Previous Article  |  Next Article »

OSA is a member of CrossRef.

CrossCheck Deposited