Border Elements Deter Shape-from-shadow via Negative Polarity, Not Motion, Not Color Annual Meeting of the Psychonomic Society, Minneapolis, MN, November 18 – 21, 2004 John M. Kennedy* & Juan Bai University of Toronto Lines, Surfaces, and Shadows Possible Segregations along a Border Shadow and Dark Line Color and Shape-from-shadow Percept Contour = reflectance change on a surface A line has two contours, one on each side We see objects by means of their surfaces, and source of illumination. A Hering line added along a shadow border may have impaired shape-from-shadow perception because of negative polarity of its contour bordering the shadow. There are different figure-ground combinations when two regions share a contour or a line (Kennedy, 1974; Peterson, 1991): Lines copying surface borders give us surface edge impressions (e. g. , the hill, the roof of the house; adapted from Kennedy, 1988). Among the list of 8 possible borders in pictures, only luminance polarity matters for shadow borders in shape-from-shadow perception, not stereo (Kennedy & Bai, 2004), and (as shown here) not color and not motion. surface borders a figure-ground percept along a contour (adapted from Rubin, 1915/1958) b figure-figure percepts along a contour But shadows without penumbras (a, adapted from Mooney, 1957) work well in showing a girl’s face, and dark lines in (b) impair perception of the face. shadow border However, a line copying a shadow border does not give us the impression of the darkness of the shadow (e. g. , the house’s shadow on the ground; adapted from Kennedy, 1988). Note that the shadowed area in (b) is darker than the illuminated region (Cavanagh & Leclerc, 1989), yet shapefrom-shadow perception is diminished. ground-figure-ground percept along a line (adapted from Kennedy et al. , 2003) figure-figure percepts along a line (adapted from Kennedy et al. , 2003) a This favors a perceptual theory (see Cavanagh & Leclerc, 1989) over figure-ground attention in explaining Hering’s ringed shadow effect. b A red line does not prevent shape-from-shadow perception as long as the luminance goes dark-to-light from shadow to line and to non-shadow (a). In comparison, the same red line blocks shape-from-shadow perception if it is darker than the shadowed region and hence creates negative border polarity (b). Motion and Shape-from-shadow Percept Border-polarity Hypothesis Kinds of Borders Hering’s Ringed Shadow figure-groundfigure percepts along a line There are 8 kinds of borders that can cause perception of surfaces, combined from luminance/spectral, monocular/binocular, and static/moving borders (Kennedy, Juricevic, & Bai, 2003). penumbra a c But lines in (c, adapted from Kennedy, 1993) do not give us the impression of the dark shadow and the elderly face. Hering, E. (1964). Outlines of a Theory of the Light Sense (L. M. Hurvich & D. Jameson, Trans. , pp. 8). Cambridge, MA: Harvard University Press. (Original work published 1874). Kennedy, J. M. (1974). A Psychology of Picture Perception. San Francisco: Jossey-Bass. Kennedy, J. M. (1988). Paper presented at the Psychonomics Society Conference, Chicago, IL. Kennedy, J. M. , Juricevic, I. , & Bai, J. (2003). In H. Hecht, R. Schwartz, & M. Atherton (Eds. ), Looking into Pictures: An Interdisciplinary Approach to Pictorial Space (pp. 321 -354). Cambridge, MA: MIT Press. Mooney, C. M. (1957). Canadian Journal of Psychology, 11, 219 -226. Peterson, M. A. (1999). Journal of Experimental Psychology: Human Perception and Performance, 25, 276 -286. b c a b c d Apparent shadow-border motion resulting from rapid alternations of (a) and (b) allows shape-from-shadow perception of the elderly face, possibly due to positive border polarity. This is in accordance with what the border-polarity hypothesis predicts! The direction of luminance change can be called ‘polarity. ’ b Goldstein, B. (2002). Sensation and Perception (6 th ed. ). Pacific Grove, Calif. : Wadsworth-Thomson Learning. Kennedy, J. M. , & Bai, J. (2004). Perception, 33, 653 -665. Both negative (a) and Hering-dark-line (b) have negative contours (i. e. , light-to-dark from shadow to non-shadow), and both fail to show the elderly face in positive (c). a Cavanagh, P. , & Leclerc, Y. G. (1989). Journal of Experimental Psychology: Human Perception and Performance, 15, 3 -27. Kennedy, J. M. & Bai, J. (2000). Perception, 29, 399 -407. a Hering (1874/1964) said a shadow with a borderline looks like a stain. He contended that for a cast shadow (a, after Goldstein, 2002), shadow perception fails in (b) because the penumbra is lost to the dark line surrounding the umbra. References Kennedy, J. M. (1993). Drawing and the Blind. New Haven: Yale Press. Shadow borders are luminance borders that may convey information about surface shapes, e. g. , the elderly face in (a, adapted from Mooney, 1957), which resembles the photo in (b). b Conclusions It may be the dark line’s negative contour bordering the shadow in (b) that diminishes shape-from-shadow perception – a ‘borderpolarity’ hypothesis that was supported by Kennedy and Bai (2000, 2004). By contrast: The same shadow-border motion resulting from alternations of negatives (c & d) does not give us the impression of the same face, possibly due to negative border polarity. This is again in accordance with what border-polarity predicts! Rubin, E. (1958). In D. C. Beardslee (Eds. ), Readings in Perception (pp. 194203). Princeton, NJ: Van Nostrand. (Original work published 1915). Acknowledgements The authors would like to extend special thanks to Humera Iqbal, Dr. Allison Sekuler, Dr. Paul Muter, Dr. Mark Schmuckler, Dr. Hiroshi Ono, and Igor Juricevic. We dedicate this poster to Rudolf Arnheim in honor of his 100 th birthday, July 15, 2004. * kennedy@utsc. utoronto. ca
Скачать презентацию Border Elements Deter Shape-from-shadow via Negative Polarity Not
97d9f9ce5776b2838fc102e10b6f9b83.ppt