This article was downloaded by: [University of Glasgow] On: 06 January 2015, At: 22:53 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
The Journal of General Psychology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/vgen20
Primary Depth Cues and Background Pattern in the Portrayal of Slant A. H. Reinhardt-Rutland
a
a
Department of Psychology , University of Ulster at Jordanstown , Northern Ireland Published online: 06 Jul 2010.
To cite this article: A. H. Reinhardt-Rutland (1992) Primary Depth Cues and Background Pattern in the Portrayal of Slant, The Journal of General Psychology, 119:1, 29-35, DOI: 10.1080/00221309.1992.9921155 To link to this article: http://dx.doi.org/10.1080/00221309.1992.9921155
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
Downloaded by [University of Glasgow] at 22:53 06 January 2015
This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
The Journal of General Psychology, I19( I ), 29-35
Downloaded by [University of Glasgow] at 22:53 06 January 2015
Primary Depth Cues and Background Pattern in the Portrayal of Slant A. H. REINHARDT-RUTLAND Department of Psychology Universiry of Ulster at Jordanstown Northern Ireland
ABSTRACT. A rectangularity postulate has been used in algorithms for the purpose of interpreting two-dimensional representations of rectilinear objects. This rectangularity postulate may affect the perception of true surfaces. In this study, rectangular surfaces and trapezoidal surfaces-the latter simulating the horizontal slant-in-depth of the rectangular surfaces-were viewed under static-monocular, movingmonocular, and static-binocular conditions, both with and without a background pattern. The static-binocular condition elicited the greatest number of accurate responses. The moving-monocular condition did not elicit significantly more accurate responses than the static-monocular viewing condition did. The effect of background pattern was insignificant. These results were unexpected in terms of ecological validity and (regarding moving-monocular viewing) because of the importance of the role of relative visual motion in the detection of object motion. However, the results are consistent with the perception of depth separation of two discrete objects.
KNOWING THAT A SURFACE IS RECTANGULAR may help a person to determine its apparent orientation-in-depth. For example, a rectangular surface slanted horizontally to the frontal plane comprises in its vertical edges differences of visual size and height-in-the-visual-field,both of which are recognized secondary or inferential cues to relative depth (Foley, 1978). The rectangularity postulate has been useful in the computational interpretation of scenes represented two dimensionally (Cowie, 1987, 1988; Mackworth, 1973). In Cowie’s algorithm, the rectangularity postulate is used to assign a The author was a joint recipient of grant GIRE 88097 of the Science and Engineering Research Council of the United Kingdom. Address correspondence to A . H . Reinhardt-Rutland, Room 17505, Department of Psychology, University of Ulster at Jordanstown, Shore Road, Newtownabbey, Co. Antrim, BT37 OQB, Northern Ireland. 29
Downloaded by [University of Glasgow] at 22:53 06 January 2015
30
The Journal of General Psychology
layout to rectilinear objects that are represented in line drawings. However, such algorithms model a restricted aspect of perception. To ensure the general applicability of the rectangularity postulate to perception of relative depth, the rectangularity postulate should be investigated in three-dimensional stimuli and in the context of primary cues due to observer motion. For example, motion parallax makes use of the relationship between visual motion and distance (Ono, Rivest, & Ono, 1986; Rogers & Graham, 1982) and, due to binocularity, binocular disparity makes use of the spatial separation of the eyes (Foley, 1978). Gehringer and Engel (1986) tested an assertion in Gibson’s (1979) treatise on the “ecological” approach to perception-that the primary cues inherent in an Ames Room stimulus should override the secondary cues. The stimulus was constructed from trapezoidal and rectangular surfaces but appeared to be the interior of a normal cuboidal room (Ittleson, 1952). Unrestrained viewing that involved both motion and binocularity failed to destroy the illusion, thus refuting Gibson’s treatise, and moving-monocular viewing was particularly ineffective. In 1990 I isolated the rectangularity postulate; observers judged the horizontal orientation-in-depth of rectangular and trapezoidal surfaces. Although binocular viewing elicited more correct answers than were elicited in Gehringer and Engel’s study, the results of the two studies were otherwise rather similar. In the present study, I extended this evidence by comparing the detection of slant-in-depth of isolated surfaces with and without a background pattern. The effect of background pattern was examined in two studies (Fems, 1972; Hell & Freeman, 1977) that involved observer motion, but background pattern did not improve depth perception in either study. In Hell and Freeman’s study, the observers determined the depth separation of two objects, and in Ferris’s study, the observers estimated the distance of a single object. Although the latter study may seem irrelevant in this context because it concerned absolute depth perception, the object in the study was viewed through an aperture in a screen, and the distance of the object might have been determined in reference to that of the aperture, had the observer been aware of it. A major difference between the previously mentioned studies and the present study is the complexity of the visual information produced by observer motion. Whereas two discrete stimuli at different distances elicit different rates of visual motion, a surface undergoes a perspective transformation that includes a variety of visual motions associated with its edges. Hence, the perceived relative depth of an object does not necessarily extrapolate to its perceived surface slant. Background pattern might conceivably enhance relative depth perception because the existence of a background pattern is more ecologically valid; an isolated surface in the visual field is unusual outside the laboratory. Also, relative visual motion is enhanced, so enhancing the cue of motion parallax.
Downloaded by [University of Glasgow] at 22:53 06 January 2015
Reinhardt-Rutland
31
The importance of relative visual motion in the interpretation of the retinal image (Nakayama, 1985; Reinhardt-Rutland, 1987, 1988) is known because of the existence of a number of phenomena that involve object motion perception. It is much easier to detect a lone object's motion if a static object is also present (Shaffer & Wallach, 1966; Wertheim & Niessen, 1986), and perceived object velocity is influenced by the velocities of surrounding objects (Loomis & Nakayama, 1973; Reinhardt-Rutland, 1991). Data from physiology (Allman, Miezen, & McGuinness, 1985;Mandl, 1985) corroborate these results. The visual system maintains an accurate coding of visual motion, but not of distal stimulation (McKee & Welch, 1989), implying that there may be considerable commonality of processing in the perception of object motion and depth based on observer motion.
Method Stimuli
The stimuli were viewed one at a time along a table top in a light-tight cubicle. The stimuli were made of matte white card and supported at the back so that they were in a vertical position. Three of the stimuli were rectangular (180 mm wide x 100 mm long) and viewed with a horizontal slant of 30", 45", or 60" to the frontal plane. The left vertical edge was closer to the observer than the right vertical edge. The remaining three stimuli were frontal but simulated the retinal images of the rectangular stimuli when viewed monocularly from the observer's chin-rest. They were thus trapezoidal, with reduced horizontal extent and reduced right vertical edge. Their dimensions were, respectively, 156 mm and 89 mm (simulating 30" slant), 127 mm and 85 mm (simulating 45" slant), and 90 mm and 83 mm (simulating 60" slant). The left edge of each stimulus was 75 cm from the observer. In one of two background conditions, I minimized the background information by painting the area surrounding the stimuli matte black. The second background condition had a row of seven vertical stripes (100 mm x 3 mm) regularly spaced horizontally over 250 mm. The stripes were visible on both sides of the stimulus, in a frontal plane 160 mm behind the left edge of the stimulus. The latter distance allowed for the increased depth of the right edge in rectangular stimuli. The stripes were made of orange retroreflective paper mounted on a rectangular sheet of transparent perspex. Although their luminance was similar to the stimuli's, their color made them easily discernible from stimuli. Two Philips TLGW08 8-watt ultraviolet tubes above the observer provided illumination. The stimulus luminance was about 0.2 cd/m2(Hagner S1 photometer). The luminance of the black areas was below measurable limits.
32
The Journal of General Psychology
Subjects
The subjects were 36 male and female undergraduates aged 19 to 30 with (by self-report) normal and uncorrected vision.
Downloaded by [University of Glasgow] at 22:53 06 January 2015
Design and Procedure
The subjects were tested individually. For half the subjects the background pattern was present during testing, and for half it was not. All the subjects viewed the stimuli monocularly with (a) head static on a chin-rest (SM); and (b) self-initiated, side-to-side head motion of 15 cm in range (MM), as well as binocularly with head static on a chin-rest (SB). In the SM and MM conditions, the subjects covered their left eye with their left hand (right-eye viewing was arbitrary because different techniques for determining eye dominance produce different results [e.g., Coren & Kaplan, 1973; Wade, 19761). In the MM condition, the subject moved his or her head to the fullest extent possible within the restraints provided, at any rate that seemed to maximize accurate responses. These instructions, in conjunction with the chosen head-motion extent, have more than adequately promoted depth perception in other displays; indeed, Rogers and Graham (1982) and Ono et al. (1986) reported that any motion promoted strong depth perception in their displays. The subjects, who were encouraged to take their time in responding, made a forced binary response of “slanted” or “facing” to each stimulus. Between trials the subjects covered their eyes while the next stimulus was being set up. The order of stimulus presentation and viewing conditions was randomized. Because each observer viewed each stimulus in each viewing condition, a maximum of six correct responses was possible for each viewing condition. Guessing would yield about three correct responses in each viewing condition.
Results and Discussion Without background stripes, the number of correct responses was SM, M = 3.2, SD = 1.2; MM, M = 3.7, SD = 1.8; and SB, M = 4.9, SD = 1.2. With background stripes, the number of correct responses was SM, M = 3.0, SD = 1.1; MM, M = 3.9, SD = 1.8; and SB, M = 5.2, SD = 0.8. The results of a two-way analysis of variance (ANOVA) indicated a significant difference across viewing conditions, F(2,60) = 32.79, p C 3 0 0 5 , but, although the presence of a background pattern was associated with slightly higher means in the MM and SB conditions, its effect was far from significant, F(l, 60) = 0.75, p = .75. The interaction between the variables was also not significant, F(2, 60) = 0.58, p = .57. A pairwise comparison of viewing conditions (Scheff6) indicated that SM and MM viewing ( p < .01
Downloaded by [University of Glasgow] at 22:53 06 January 2015
Reinhardt-Rutland
33
for both) were each significantly different from SB viewing, but MM viewing was not significantly different from SM viewing. The pattern of errors supports the view that the rectangularity postulate played an important role in determining the subjects’ responses: 67% of the errors were due to facing stimuli being labeled as slanted. The proportions of errors for the three simulated and real orientations were 35% (30”), 32% (45”), and 34% (60”), indicating little difference in this variable. Regarding viewing conditions, the findings of the present study are similar to those of my 1990 study and Gehringer and Engel’s (1986) study: Staticmonocular viewing was the least accurate condition, moving-monocular viewing was a little better, and static-binocular viewing was much more accurate. The results of this study and those of my 1990 study indicated more accurate responses in the static-binocular viewing condition than did the results of Gehringer and Engel’s (1986) study, in which approximately 60% of the maximum illusory effect remained. Incidentally, the procedure in Gehringer and Engel’s study always allowed some motion. Motion cues have been ineffective in other contexts of relative depth perception. For example, observer motion failed to elicit the correct answers regarding the order-in-depth of objects at different distances when visual size and height in the visual field conveyed competing information to the observer (Eriksson, 1972 a,b). The effect of background pattern was consistent with the results of Hell and Freeman’s (1977) and Ferris’s (1972) studies: A seemingly ecological intervention fails to elicit significantly more accurate responses. Presumably, detection of slant in this experiment is heavily reliant on information that is internal to the surface. Similarly, two objects (Hell & Freeman, 1977) or an object relative to an aperture (Ferris, 1972) may be regarded more or less as integral “systems” as far as depth perception is concerned. Most of the evidence supporting the importance of relative visual motion has concerned the perception of object motion. Although there is reason to extrapolate some of this evidence to perception that involves observer motion, not all the evidence is supportive. Motion aftereffect is perceived motion of a static pattern that is observed after a person has viewed actual motion, and probably reflects sensory processing of the actual motion. Two conditions must be met if motion aftereffect is to be observed. First, the motion must be relative; no motion aftereffect is observed after a uniform motion is viewed in the absence of a patterned background (Day & Strelow, 1971; Naatanen, 1973). Second, no motion aftereffect is observed if the actual motion can be attributed to the observer’s movement with respect to static stimuli, although the same motion produces an aftereffect if the observer is motionless (Denton, 1977; Harris, Morgan, & Still, 1981). Other evidence derived from a number of motion phenomena, including motion aftereffects, concerns postulates about objects. Motion is preferentially assigned to relatively small objects that are somewhat central in the
Downloaded by [University of Glasgow] at 22:53 06 January 2015
34
The Journal of General Psychology
visual field (Reinhardt-Rutland, 1987, 1988). However, such postulates must be ovemdden if, as generally applies in depth perception, such an object is perceived as static. Apparently, conclusions inferred from object motion do not necessarily apply to distance perception based on an observer’s motion. One can conclude that the rectangularity postulate is useful in contexts other than two-dimensional representation (Cowie, 1987, 1988; Mackworth, 1973). Although binocular viewing elicited relative accuracy in this study, binocular cues are effective only if they are viewed at fairly close range (Foley, 1978). The rectangularity postulate may well be important for binocular viewing at greater distances than that used in the present study, as suggested by the relative ineffectiveness of binocular viewing in Gehringer and Engel’s (1986) study. Another reason for Gehringer and Engel’s relatively poor results for binocular viewing is that the Ames stimuli may reflect other postulates in addition to the rectangularity postulate-plausibly perpendicularity of surfacejunctions. If algorithms for scene analysis are extrapolated to three-dimensional perception, such additional postulates might be useful. REFERENCES Allman, J., Miezin, F., & McGuinness, E. (1985). Direction-and velocity-specific responses from beyond the classical receptive field in the middle femporal visual area (MT). Perception, 14, 105-126. Coren, S., & Kaplan, C. P. (1973). Patterns of ocular dominance. American Journal of Optometry and Archives of the American Academy of Optometry, 50, 283-292. Cowie, R. I. D. (1987). The alternative allowed by a rectangularity postulate, and a pragmatic approach to interpreting motion. In J. Hallam & C. Mellish (Eds.), Advances in artijicial intelligence (pp. 109-1 2 1). New York: Wiley. Cowie, R. I. D. (1988). Impossible objects and the things we do first in vision. British Journal of Psychology, 79, 321-338. Day, R. H., & Strelow, E. (1971). Reduction or disappearance of visual after-effect of movement in the absence of patterned surround. Nature, 230, 55-56. Denton, G. G. (1977). Visual motion aftereffect induced by simulated rectilinear motion. Perception, 6, 71 1-718. Eriksson, E. S. (1972a). Movement parallax, anisotrophy, and relative size as determinants of space perception. Report 131: Department of Psychology, University of Uppsala, Sweden. Eriksson, E. S. (1972b). The effects of movement parallax on vertically separated stimuli at different distances. Report 1 19: Department of Psychology, University of Uppsala, Sweden. Ferris, S. H. (1972). Motion parallax and absolute distance. Journal of Experimental PSyChOlogy, 95, 258-263. Foley, J. M. (1978). Primary distance cues. In R. Held, H. W. Leibowitz, & H. L. Teuber (Eds.), Handbook of sensory physiology (Vol. 8, pp. 181-213). Berlin: Springer-Verlag. Gehringer, W. L., & Engel, E. (1986). Effect of ecological viewing conditions on the Ames’ distorted room illusion. Journal of Experimental Psychology: Human Perception and Performance, 12, 181-185.
Downloaded by [University of Glasgow] at 22:53 06 January 2015
Gibson, J. J. (1979). The ecological approach to visual perception. Boston: Houghton Mifflin. Harris,L. R., Morgan, M. J., & Still, A. W. (1981). Moving and the motion aftereffect. Nature, 293, 139-141. Hell, W., & Freeman, R. B. (1977). Detectability of motion as a factor in depth perception by monocular movement parallax. Perception and Psychophysics, 22, 526-530. Ittleson, W. (1952). The Ames demonstrations in perception. Princeton, NJ: Princeton University Press. Loomis, J. M., & Nakayama, K. (1973). A velocity analogue of brightness contrast. Perception, 2 , 425-428. Mackworth, A. K. (1973). Interpreting pictures of polyhedral scenes. Artificial lntelligence, 4 , 121-137. Mandl, G. (1985). Responses of visual cells in cat superior colliculus to relative pattern movement. Vision Research, 25, 267-281. McKee, S . P., & Welch, L. (1989). Is there a constancy for velocity? Vision Research, 29, 553-561. Naatanen, R. (1973). Effect of stimulus background on waterfall illusion. Perceptual and Motor Skills, 37, 762. Nakayama, K. (1985). Biological image motion processing: A review. Vision Research, 25, 625-660. Ono, M. E., Rivest, J., & Ono, H. (1986). Depth perception as a function of motion parallax and absolute-distance information. Journal of Experimental Psychology: Human Perception and Performance, 12, 33 1-337. Reinhardt-Rutland, A. H. (1987). Aftereffect of visual movement-The role of relative movement: A review. Current Psychological Research and Reviews, 6, 275288. Reinhardt-Rutland, A. H. (1988). Induced movement in the visual modality: An overview. Psychology Bulletin, 103 52-71. Reinhardt-Rutland, A. H. (1990). Detecting orientation of a surface: The rectangularity postulate and primary depth cues. Journal of General Psychology, 117, 391401. Reinhardt-Rutland, A. H. (1991). Induced rotary movement during eye movements: Displays with unequal spacing of pattern. Journal of General Psychology, 118, 129-137. Rogers, B., & Graham, M. (1982). Similarities between motion parallax and stereopsis in human depth perception. Vision Research, 22, 261-270. ShaEer, 0..& Wallach, H. (1966). Extent-of-motion thresholds under subjectrelative and object-relative conditions. Perception and Psychophysics, I , 447451. Wade, N. J. (1976). On interocular transfer of the movement aftereffect in individuals with and without normal binocular vision. Perception, 5 , 113-1 18. Wertheim, A. H., & Niessen, M. W. (1986). The perception of relative motion between objects during pursuit eye movements. Perception, 15, A9.
Received September 10, I991
The Journal of General Psychology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/vgen20
Primary Depth Cues and Background Pattern in the Portrayal of Slant A. H. Reinhardt-Rutland
a
a
Department of Psychology , University of Ulster at Jordanstown , Northern Ireland Published online: 06 Jul 2010.
To cite this article: A. H. Reinhardt-Rutland (1992) Primary Depth Cues and Background Pattern in the Portrayal of Slant, The Journal of General Psychology, 119:1, 29-35, DOI: 10.1080/00221309.1992.9921155 To link to this article: http://dx.doi.org/10.1080/00221309.1992.9921155
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
Downloaded by [University of Glasgow] at 22:53 06 January 2015
This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
The Journal of General Psychology, I19( I ), 29-35
Downloaded by [University of Glasgow] at 22:53 06 January 2015
Primary Depth Cues and Background Pattern in the Portrayal of Slant A. H. REINHARDT-RUTLAND Department of Psychology Universiry of Ulster at Jordanstown Northern Ireland
ABSTRACT. A rectangularity postulate has been used in algorithms for the purpose of interpreting two-dimensional representations of rectilinear objects. This rectangularity postulate may affect the perception of true surfaces. In this study, rectangular surfaces and trapezoidal surfaces-the latter simulating the horizontal slant-in-depth of the rectangular surfaces-were viewed under static-monocular, movingmonocular, and static-binocular conditions, both with and without a background pattern. The static-binocular condition elicited the greatest number of accurate responses. The moving-monocular condition did not elicit significantly more accurate responses than the static-monocular viewing condition did. The effect of background pattern was insignificant. These results were unexpected in terms of ecological validity and (regarding moving-monocular viewing) because of the importance of the role of relative visual motion in the detection of object motion. However, the results are consistent with the perception of depth separation of two discrete objects.
KNOWING THAT A SURFACE IS RECTANGULAR may help a person to determine its apparent orientation-in-depth. For example, a rectangular surface slanted horizontally to the frontal plane comprises in its vertical edges differences of visual size and height-in-the-visual-field,both of which are recognized secondary or inferential cues to relative depth (Foley, 1978). The rectangularity postulate has been useful in the computational interpretation of scenes represented two dimensionally (Cowie, 1987, 1988; Mackworth, 1973). In Cowie’s algorithm, the rectangularity postulate is used to assign a The author was a joint recipient of grant GIRE 88097 of the Science and Engineering Research Council of the United Kingdom. Address correspondence to A . H . Reinhardt-Rutland, Room 17505, Department of Psychology, University of Ulster at Jordanstown, Shore Road, Newtownabbey, Co. Antrim, BT37 OQB, Northern Ireland. 29
Downloaded by [University of Glasgow] at 22:53 06 January 2015
30
The Journal of General Psychology
layout to rectilinear objects that are represented in line drawings. However, such algorithms model a restricted aspect of perception. To ensure the general applicability of the rectangularity postulate to perception of relative depth, the rectangularity postulate should be investigated in three-dimensional stimuli and in the context of primary cues due to observer motion. For example, motion parallax makes use of the relationship between visual motion and distance (Ono, Rivest, & Ono, 1986; Rogers & Graham, 1982) and, due to binocularity, binocular disparity makes use of the spatial separation of the eyes (Foley, 1978). Gehringer and Engel (1986) tested an assertion in Gibson’s (1979) treatise on the “ecological” approach to perception-that the primary cues inherent in an Ames Room stimulus should override the secondary cues. The stimulus was constructed from trapezoidal and rectangular surfaces but appeared to be the interior of a normal cuboidal room (Ittleson, 1952). Unrestrained viewing that involved both motion and binocularity failed to destroy the illusion, thus refuting Gibson’s treatise, and moving-monocular viewing was particularly ineffective. In 1990 I isolated the rectangularity postulate; observers judged the horizontal orientation-in-depth of rectangular and trapezoidal surfaces. Although binocular viewing elicited more correct answers than were elicited in Gehringer and Engel’s study, the results of the two studies were otherwise rather similar. In the present study, I extended this evidence by comparing the detection of slant-in-depth of isolated surfaces with and without a background pattern. The effect of background pattern was examined in two studies (Fems, 1972; Hell & Freeman, 1977) that involved observer motion, but background pattern did not improve depth perception in either study. In Hell and Freeman’s study, the observers determined the depth separation of two objects, and in Ferris’s study, the observers estimated the distance of a single object. Although the latter study may seem irrelevant in this context because it concerned absolute depth perception, the object in the study was viewed through an aperture in a screen, and the distance of the object might have been determined in reference to that of the aperture, had the observer been aware of it. A major difference between the previously mentioned studies and the present study is the complexity of the visual information produced by observer motion. Whereas two discrete stimuli at different distances elicit different rates of visual motion, a surface undergoes a perspective transformation that includes a variety of visual motions associated with its edges. Hence, the perceived relative depth of an object does not necessarily extrapolate to its perceived surface slant. Background pattern might conceivably enhance relative depth perception because the existence of a background pattern is more ecologically valid; an isolated surface in the visual field is unusual outside the laboratory. Also, relative visual motion is enhanced, so enhancing the cue of motion parallax.
Downloaded by [University of Glasgow] at 22:53 06 January 2015
Reinhardt-Rutland
31
The importance of relative visual motion in the interpretation of the retinal image (Nakayama, 1985; Reinhardt-Rutland, 1987, 1988) is known because of the existence of a number of phenomena that involve object motion perception. It is much easier to detect a lone object's motion if a static object is also present (Shaffer & Wallach, 1966; Wertheim & Niessen, 1986), and perceived object velocity is influenced by the velocities of surrounding objects (Loomis & Nakayama, 1973; Reinhardt-Rutland, 1991). Data from physiology (Allman, Miezen, & McGuinness, 1985;Mandl, 1985) corroborate these results. The visual system maintains an accurate coding of visual motion, but not of distal stimulation (McKee & Welch, 1989), implying that there may be considerable commonality of processing in the perception of object motion and depth based on observer motion.
Method Stimuli
The stimuli were viewed one at a time along a table top in a light-tight cubicle. The stimuli were made of matte white card and supported at the back so that they were in a vertical position. Three of the stimuli were rectangular (180 mm wide x 100 mm long) and viewed with a horizontal slant of 30", 45", or 60" to the frontal plane. The left vertical edge was closer to the observer than the right vertical edge. The remaining three stimuli were frontal but simulated the retinal images of the rectangular stimuli when viewed monocularly from the observer's chin-rest. They were thus trapezoidal, with reduced horizontal extent and reduced right vertical edge. Their dimensions were, respectively, 156 mm and 89 mm (simulating 30" slant), 127 mm and 85 mm (simulating 45" slant), and 90 mm and 83 mm (simulating 60" slant). The left edge of each stimulus was 75 cm from the observer. In one of two background conditions, I minimized the background information by painting the area surrounding the stimuli matte black. The second background condition had a row of seven vertical stripes (100 mm x 3 mm) regularly spaced horizontally over 250 mm. The stripes were visible on both sides of the stimulus, in a frontal plane 160 mm behind the left edge of the stimulus. The latter distance allowed for the increased depth of the right edge in rectangular stimuli. The stripes were made of orange retroreflective paper mounted on a rectangular sheet of transparent perspex. Although their luminance was similar to the stimuli's, their color made them easily discernible from stimuli. Two Philips TLGW08 8-watt ultraviolet tubes above the observer provided illumination. The stimulus luminance was about 0.2 cd/m2(Hagner S1 photometer). The luminance of the black areas was below measurable limits.
32
The Journal of General Psychology
Subjects
The subjects were 36 male and female undergraduates aged 19 to 30 with (by self-report) normal and uncorrected vision.
Downloaded by [University of Glasgow] at 22:53 06 January 2015
Design and Procedure
The subjects were tested individually. For half the subjects the background pattern was present during testing, and for half it was not. All the subjects viewed the stimuli monocularly with (a) head static on a chin-rest (SM); and (b) self-initiated, side-to-side head motion of 15 cm in range (MM), as well as binocularly with head static on a chin-rest (SB). In the SM and MM conditions, the subjects covered their left eye with their left hand (right-eye viewing was arbitrary because different techniques for determining eye dominance produce different results [e.g., Coren & Kaplan, 1973; Wade, 19761). In the MM condition, the subject moved his or her head to the fullest extent possible within the restraints provided, at any rate that seemed to maximize accurate responses. These instructions, in conjunction with the chosen head-motion extent, have more than adequately promoted depth perception in other displays; indeed, Rogers and Graham (1982) and Ono et al. (1986) reported that any motion promoted strong depth perception in their displays. The subjects, who were encouraged to take their time in responding, made a forced binary response of “slanted” or “facing” to each stimulus. Between trials the subjects covered their eyes while the next stimulus was being set up. The order of stimulus presentation and viewing conditions was randomized. Because each observer viewed each stimulus in each viewing condition, a maximum of six correct responses was possible for each viewing condition. Guessing would yield about three correct responses in each viewing condition.
Results and Discussion Without background stripes, the number of correct responses was SM, M = 3.2, SD = 1.2; MM, M = 3.7, SD = 1.8; and SB, M = 4.9, SD = 1.2. With background stripes, the number of correct responses was SM, M = 3.0, SD = 1.1; MM, M = 3.9, SD = 1.8; and SB, M = 5.2, SD = 0.8. The results of a two-way analysis of variance (ANOVA) indicated a significant difference across viewing conditions, F(2,60) = 32.79, p C 3 0 0 5 , but, although the presence of a background pattern was associated with slightly higher means in the MM and SB conditions, its effect was far from significant, F(l, 60) = 0.75, p = .75. The interaction between the variables was also not significant, F(2, 60) = 0.58, p = .57. A pairwise comparison of viewing conditions (Scheff6) indicated that SM and MM viewing ( p < .01
Downloaded by [University of Glasgow] at 22:53 06 January 2015
Reinhardt-Rutland
33
for both) were each significantly different from SB viewing, but MM viewing was not significantly different from SM viewing. The pattern of errors supports the view that the rectangularity postulate played an important role in determining the subjects’ responses: 67% of the errors were due to facing stimuli being labeled as slanted. The proportions of errors for the three simulated and real orientations were 35% (30”), 32% (45”), and 34% (60”), indicating little difference in this variable. Regarding viewing conditions, the findings of the present study are similar to those of my 1990 study and Gehringer and Engel’s (1986) study: Staticmonocular viewing was the least accurate condition, moving-monocular viewing was a little better, and static-binocular viewing was much more accurate. The results of this study and those of my 1990 study indicated more accurate responses in the static-binocular viewing condition than did the results of Gehringer and Engel’s (1986) study, in which approximately 60% of the maximum illusory effect remained. Incidentally, the procedure in Gehringer and Engel’s study always allowed some motion. Motion cues have been ineffective in other contexts of relative depth perception. For example, observer motion failed to elicit the correct answers regarding the order-in-depth of objects at different distances when visual size and height in the visual field conveyed competing information to the observer (Eriksson, 1972 a,b). The effect of background pattern was consistent with the results of Hell and Freeman’s (1977) and Ferris’s (1972) studies: A seemingly ecological intervention fails to elicit significantly more accurate responses. Presumably, detection of slant in this experiment is heavily reliant on information that is internal to the surface. Similarly, two objects (Hell & Freeman, 1977) or an object relative to an aperture (Ferris, 1972) may be regarded more or less as integral “systems” as far as depth perception is concerned. Most of the evidence supporting the importance of relative visual motion has concerned the perception of object motion. Although there is reason to extrapolate some of this evidence to perception that involves observer motion, not all the evidence is supportive. Motion aftereffect is perceived motion of a static pattern that is observed after a person has viewed actual motion, and probably reflects sensory processing of the actual motion. Two conditions must be met if motion aftereffect is to be observed. First, the motion must be relative; no motion aftereffect is observed after a uniform motion is viewed in the absence of a patterned background (Day & Strelow, 1971; Naatanen, 1973). Second, no motion aftereffect is observed if the actual motion can be attributed to the observer’s movement with respect to static stimuli, although the same motion produces an aftereffect if the observer is motionless (Denton, 1977; Harris, Morgan, & Still, 1981). Other evidence derived from a number of motion phenomena, including motion aftereffects, concerns postulates about objects. Motion is preferentially assigned to relatively small objects that are somewhat central in the
Downloaded by [University of Glasgow] at 22:53 06 January 2015
34
The Journal of General Psychology
visual field (Reinhardt-Rutland, 1987, 1988). However, such postulates must be ovemdden if, as generally applies in depth perception, such an object is perceived as static. Apparently, conclusions inferred from object motion do not necessarily apply to distance perception based on an observer’s motion. One can conclude that the rectangularity postulate is useful in contexts other than two-dimensional representation (Cowie, 1987, 1988; Mackworth, 1973). Although binocular viewing elicited relative accuracy in this study, binocular cues are effective only if they are viewed at fairly close range (Foley, 1978). The rectangularity postulate may well be important for binocular viewing at greater distances than that used in the present study, as suggested by the relative ineffectiveness of binocular viewing in Gehringer and Engel’s (1986) study. Another reason for Gehringer and Engel’s relatively poor results for binocular viewing is that the Ames stimuli may reflect other postulates in addition to the rectangularity postulate-plausibly perpendicularity of surfacejunctions. If algorithms for scene analysis are extrapolated to three-dimensional perception, such additional postulates might be useful. REFERENCES Allman, J., Miezin, F., & McGuinness, E. (1985). Direction-and velocity-specific responses from beyond the classical receptive field in the middle femporal visual area (MT). Perception, 14, 105-126. Coren, S., & Kaplan, C. P. (1973). Patterns of ocular dominance. American Journal of Optometry and Archives of the American Academy of Optometry, 50, 283-292. Cowie, R. I. D. (1987). The alternative allowed by a rectangularity postulate, and a pragmatic approach to interpreting motion. In J. Hallam & C. Mellish (Eds.), Advances in artijicial intelligence (pp. 109-1 2 1). New York: Wiley. Cowie, R. I. D. (1988). Impossible objects and the things we do first in vision. British Journal of Psychology, 79, 321-338. Day, R. H., & Strelow, E. (1971). Reduction or disappearance of visual after-effect of movement in the absence of patterned surround. Nature, 230, 55-56. Denton, G. G. (1977). Visual motion aftereffect induced by simulated rectilinear motion. Perception, 6, 71 1-718. Eriksson, E. S. (1972a). Movement parallax, anisotrophy, and relative size as determinants of space perception. Report 131: Department of Psychology, University of Uppsala, Sweden. Eriksson, E. S. (1972b). The effects of movement parallax on vertically separated stimuli at different distances. Report 1 19: Department of Psychology, University of Uppsala, Sweden. Ferris, S. H. (1972). Motion parallax and absolute distance. Journal of Experimental PSyChOlogy, 95, 258-263. Foley, J. M. (1978). Primary distance cues. In R. Held, H. W. Leibowitz, & H. L. Teuber (Eds.), Handbook of sensory physiology (Vol. 8, pp. 181-213). Berlin: Springer-Verlag. Gehringer, W. L., & Engel, E. (1986). Effect of ecological viewing conditions on the Ames’ distorted room illusion. Journal of Experimental Psychology: Human Perception and Performance, 12, 181-185.
Downloaded by [University of Glasgow] at 22:53 06 January 2015
Gibson, J. J. (1979). The ecological approach to visual perception. Boston: Houghton Mifflin. Harris,L. R., Morgan, M. J., & Still, A. W. (1981). Moving and the motion aftereffect. Nature, 293, 139-141. Hell, W., & Freeman, R. B. (1977). Detectability of motion as a factor in depth perception by monocular movement parallax. Perception and Psychophysics, 22, 526-530. Ittleson, W. (1952). The Ames demonstrations in perception. Princeton, NJ: Princeton University Press. Loomis, J. M., & Nakayama, K. (1973). A velocity analogue of brightness contrast. Perception, 2 , 425-428. Mackworth, A. K. (1973). Interpreting pictures of polyhedral scenes. Artificial lntelligence, 4 , 121-137. Mandl, G. (1985). Responses of visual cells in cat superior colliculus to relative pattern movement. Vision Research, 25, 267-281. McKee, S . P., & Welch, L. (1989). Is there a constancy for velocity? Vision Research, 29, 553-561. Naatanen, R. (1973). Effect of stimulus background on waterfall illusion. Perceptual and Motor Skills, 37, 762. Nakayama, K. (1985). Biological image motion processing: A review. Vision Research, 25, 625-660. Ono, M. E., Rivest, J., & Ono, H. (1986). Depth perception as a function of motion parallax and absolute-distance information. Journal of Experimental Psychology: Human Perception and Performance, 12, 33 1-337. Reinhardt-Rutland, A. H. (1987). Aftereffect of visual movement-The role of relative movement: A review. Current Psychological Research and Reviews, 6, 275288. Reinhardt-Rutland, A. H. (1988). Induced movement in the visual modality: An overview. Psychology Bulletin, 103 52-71. Reinhardt-Rutland, A. H. (1990). Detecting orientation of a surface: The rectangularity postulate and primary depth cues. Journal of General Psychology, 117, 391401. Reinhardt-Rutland, A. H. (1991). Induced rotary movement during eye movements: Displays with unequal spacing of pattern. Journal of General Psychology, 118, 129-137. Rogers, B., & Graham, M. (1982). Similarities between motion parallax and stereopsis in human depth perception. Vision Research, 22, 261-270. ShaEer, 0..& Wallach, H. (1966). Extent-of-motion thresholds under subjectrelative and object-relative conditions. Perception and Psychophysics, I , 447451. Wade, N. J. (1976). On interocular transfer of the movement aftereffect in individuals with and without normal binocular vision. Perception, 5 , 113-1 18. Wertheim, A. H., & Niessen, M. W. (1986). The perception of relative motion between objects during pursuit eye movements. Perception, 15, A9.
Received September 10, I991
