Effects of Eye Condition and Background Illumination on Depth Perception

Abstract
Effects of Eye Conditon and Backround Illumination on Depth Perception
Subjects
Apparatus
Procedure
Results
Discussion
References

Abstract

An experiment was conducted to investigate depth perception. The phenomena of interest were eye condition (whether depth perception was better with one or two eyes) and the effects of background illumination on depth perception. Two hypotheses were tested: (1) depth perception would be better when the subject used binocular vision, and (2) depth perception would increase as the intensity of background illumination increased. The 15 subjects were required to line up vertical rods in a Howard-Dolman box, using binocular and monocular vision under the four levels of background illumination, which were 1, 3, 6, and 12 footcandles. A 2 x 4 analysis of variance found support or the first hypothesis, F(1,98) = 9.19, p <0.01, but not for the second. Results obtained for eye condition were in agreement with those obtained by Berry (1948), Chalmers (1952), and Barrett and Williamson (1966).


Effects of Eye Condition and Background Illumination on Depth Perception

Depth perception has been a topic of interest for many years. For example, the renaissance painter used depth cues to improve the quality of his art. Scientists and artists have used depth cues and the ability to confuse depth cues as a source of investigation and entertainment. Distance judgments are partially a function of how distant an object is, and the depth cues that are used to inform us of the distance of an object.

There are a number of sources of information for depth cues, these are defined here for reference: accommodation (the change in the shape of the lens in your eye as you focus on objects at different distances); convergence (the two eyes turning inward or outward as they attempt to focus on an object); retinal disparity (the difference in image cast by an object on the retinae of the eyes as the object moves closer or farther away); linear perspective (when parallel lines appear to converge as they recede into the distance); interposition (objects that are partially obscured will appear to be farther away); texture gradient (objects that are farther away appear to be smaller and closer together); atmospheric perspective (the observation that distant objects often look blurry and bluish, in contrast to nearby objects, because particles in the air will cause refractions of light rays and affect perception of the object); relative motion/motion parallax (nearby objects appear to move more rapidly in relation to our own motion, objects that are farther away tend to look like they are moving slower); relative size (the tendency to perceive an object as being the same size even as the size of its retinal image changes according to distance); relative distance (how far two objects are from each other); patterns of light and shade (opaque objects block light and produce shadows); Matlin (1988) and Rathus (1990).

The primary depth cues are: accommodation, convergence, and retinal disparity. Accommodation is a monocular cue for depth perception, while convergence and retinal disparity are binocular cues for depth. The remainder are all secondary depth cues, and are monocular cues for depth perception.

Berry (1948) conducted an experiment to examine visual perception in three situations; stereoscopic depth, real depth, and vernier or lateral displacement. The experiment used 3 male subjects. The apparatus was two black vertical rods, one above the other. The examination of depth perception was based on the summation of the angles formed by the line of sight from the subjects eyes to the test object.

From the data collected it was apparent that stereoscopic depth and real depth were quite similar for each subject, and that each was similarly affected by changes in the vertical separation between the rods. In the stereoscopic situation, the fused image of the upper rods appeared to be nearer to the observer than the similarly fused image of the lower rods. In the real depth situation, the 2 vertical rods were placed such that the upper rod was closer to the observer than the lower rod.

The vernier thresholds were quite differently affected by the factor of separation. In the vernier situation, the problem was a barely perceptible depth displacement in its relation to a barely perceptible vernier or lateral displacement. The rods in this case were arranged beside each other.

In Berry's experiment real depth and stereoscopic depth situations yielded essentially the same thresholds. It was demonstrated that the thresholds for stereoscopic depth and real depth were almost identical for all vertical separations of the test rods. However, the relation between the vernier thresholds and the depth thresholds differed as the vertical separations were varied.

There was not a binocular summative effect when the two images on the retinae were identical. However when these two images were not identical, as in real depth or stereoscopic depth the binocular fusion did result in a summative effect.

Three visual discrimination situations were presented to each of the subjects. Results for vernier tests found that using monocular vision was in general equal to those found using binocular vision. Stereoscopic and real depth discriminations could not be interpreted as merely a combination of vernier or monocular discriminations.

Chalmers (1952) conducted research to study primary monocular and binocular depth cues which enter into perception of size at distances of 100 ft. There were 5 subjects who sat in a dark tunnel, 120 ft long, with a shadow box at the end. Directly in front of the subject were eight panels, hung at different distances. Each panel had a triangular opening through which the subject could see the illuminated surface of a shadow box. Eight standard triangles were cut to different sizes so that each one at its respective distance would occupy a visual angle of 45°. Only one of the standard stimulus-triangles was in viewing position at a time.

The experiment was performed by comparing monocular and binocular determinations with a comparison triangle at 10 or 120 ft. Subjects were asked to tell the experimenter when the comparison triangle seemed to be equal in real, physical size to the standard triangle.

Monocular vision provided information about the size or distance of objects to about 25 ft away. When using binocular vision, it was possible for subjects to perceive the real, physical size and presumably the distance of objects up to 80 ft away.

Barrett and Williamson (1966) Conducted an experiment in which subjects judged the quality of depth of a three-dimensional scene. The subjects were 9 males and 6 females, ranging in age from 20 to 40. The stimuli were two model cars approximately 7 inches long and 2 3/4 in. wide. The model cars were placed on a dull black surface 22 ft in front of the subject and 5 in. apart, one behind the other but with a three-quarter overlap and sides perpendicular to the line of sight. A paired comparison technique was used to obtain the data. The scenes remained the same, but the eye-viewing combination changed. Eighteen trials were presented in a random sequence. Subjects compared and judged the scenes on a seven-point scale for quality or sensation of depth. The subjects were given instructions to keep both eyes open, but one eye or the other (or neither) would be occluded without their awareness, by using a polarizing filter on a eyepiece on the viewing console. In 70.8% of instances, the subjects judged the binocular viewed scene to be better than the monocular viewed scene. The data indicated that subjects were more confident of their response when they choose the binocular rather then the monocular viewed scene.

Based on the above review, it is hypothesized that if more depth cues are available with both eyes open, then depth perception will be better with binocular vision than with monocular vision.

Mueller and Lloyd (1948) conducted an experiment to determine the effects of different levels of illumination on visual acuity. Two subjects viewed a non-movable target (a long vertical line) and a movable target through a stereoscopic viewing device. The 2 subjects each provided three complete sets of data, each set of data consisted of 20 readings taken at each of ten intensity levels. The intensity levels were presented to the subject in order of increasing magnitude. The data indicated that stereoscopic acuity increased as light intensity increased until at high intensities, the curve approached a final limiting value.

Berry, Riggs, and Duncan (1950) conducted an experiment in which the subjects were 3 males. The experiment was an outgrowth of Berry's (1948) experiment in which functions of vernier, stereoscopic and real depth were examined. This 1950 study attempted to measure vernier and real depth thresholds using various levels of brightness of the visual field. The test objects were two vertical steel rods placed one above the other, with a 3 mm vertical separation between them. Each subject participated in a total of 24 sessions for the six levels of brightness for vernier and real depth. Berry et al. concluded that depth perception was aided more than vernier discriminations by increased levels of light.

Coules (1955) conducted an inquiry to examine the effects of brightness and depth perception. Two experiments were performed. The purpose of the first experiment was to determine the functional relationship between brightness and judgments of distance under binocular and monocular viewing conditions, and to obtain stimulus-stimulus functions for brightness and object distance. The purpose of the second experiment was to determine whether the effect of brightness on distance judgments could be generalized by varying the absolute level of brightness. There were 2 subjects, one male and one female. In general, the results of the first experiment indicated that the brighter object appeared nearer than the dimmer object.

In order to compare the influence of brightness under binocular and monocular conditions, it was necessary to keep distance constant. As the brightness increased the percentage of frequency of "nearer" responses increased, suggesting that brightness affected distance judgments independent of retinal disparity and convergence.
Inconsistent results were obtained in the second experiment between the brightness level and point of subjective equity for a given brightness ratio.

Based on review of the above, it is hypothesized that if background illumination is increased, then the ability to perceive depth will be increased.

Subjects

The subjects were 4 male and 11 female students from an Introduction to Experimental Psychology class of a medium sized university in Southern California. The subjects ages ranged from 19 to 46 years. Each subject was tested for visual acuity using the Snellen Eye Chart. Only subjects with 20/20 vision, or vision corrected to 20/20 participated in the experiment.

Apparatus

The experiment was conducted in one session in a room measuring 9.6 x 11.1 x 2.6 m, the room had 6 windows and 3 doors. The doors were closed and the blinds were drawn to exclude external light. There was a small light source to enable the experimenters to record the data.


In the front of the room there were eight tables, they measured 295 x 59 x 94 cm, and were spaced 134 cm apart. On each table there was a Howard-Dolman box, measuring 50.5 x 18.5 x 20 cm, the layout and dimensions of the Howard-Dolman box are presented in Figure 1.

figure 1

Figure 1: Three dimensional layout of the Howard-Dolman box. Measurements are in centimeters.


The front panel of the Howard-Dolman box was 18.5 x 25 cm. The top of the front panel was 5 cm higher than the main body of the box to prevent the subject from seeing the scale on top of the box and using it as a reference when aligning the rods. There was a rectangular front opening 12.5 x 7 cm which was 6 cm above the table. The strings that the subject used to align the rods were routed through two holes in the front panel. Inside the box were two vertical rods measuring 5 mm in diameter, 15 cm high, 6 cm apart, attached to strings on a pulley. There was scale on the top of the box which was used to indicate the alignment of the rods. A pointer was attached to one of the rods, such that when the strings were pulled, the pointer would move along the scale. The pointer indicated the subject's score for rod alignment. The scale had graduations ranging +100 to -100 with 0 in the center indicating perfect alignment of the rods. Behind the rods, 38 cm from the front panel, was a semi-opaque panel made of Plexiglas, with a light bulb behind it. The light bulb was attached to the Variac resistor, which was used to adjust the level of background illumination for the four levels; 1, 3, 6, and 12 footcandles.

Procedure

The experiment was run in one session, half of the participants served as subjects, the remaining half served as experimenters in each of the eight trials. All subjects were tested using all levels of background illumination and both eye conditions once. The subjects and experimenters then switched roles for an additional eight trials.

Subjects were dark adapted for 5 minutes to the experimental environment. The subjects sat across from the darkened box, then experimenters randomly selected a starting point on the scale and an illumination condition. The subjects were then instructed to align the rods. When the subjects were satisfied with the alignment, they put down the string indicating the end of that trial and the experimenter recorded the results. If the rods touched the end of the scale or there was excessive head movement by the subject, it was declared a miss-trial, and was repeated.

Results

The effects of eye condition and background illumination on depth perception are presented in Figure 2.

figure 2

Figure 2. Results of the effects of eye condition and background illumination on depth perception.



A 2 x 4 repeated analysis of variance was performed on the data collected. The results are summarized in Table 1.

Table 1
Analysis of Variance Summary
Source Sum of Squares df Mean Square F Ratio
1. Within 6147.88 14 439.13 1.39 *
2. Treatments 3330.53 7
3. A 2900.83 1 2900.83 9.19 **
4. B 157.53 3 52.51 0.17 *
5. AB 272.17 3 90.72 0.29 *
6. Residual 30925.59 98 315.57
7. Total 40404 119
* p > 0.05
** p < 0.01

Support was found for the first hypothesis,
F(1,98) = 9.19, p <0.01. This indicated that the main effect of eye condition with one or two eyes does make a significant difference on depth perception.

The second hypothesis was not supported,
F(3,98) = 0.17, p >0.05. Indicating that the level of background illumination did not make a significant difference on depth perception.


Discussion

In this experiment it was expected that binocular vision would provide better depth perception than monocular vision. This hypothesis was supported. The mean depth response was better when subjects used both eyes to line up the vertical bars in the Howard-Dolman box.

An examination of Figure 2 reveals the depth perception scores when using binocular or monocular vision. From this data, ratios of binocular to monocular depth perception were calculated. The lowest ratio was found when the background illumination was 3 footcandles, the results that depth perception trial was an average score of 11 mm using binocular vision, and an average score of 16 mm using monocular vision, for a ratio of 1:1.45. The highest ratio was when the background illumination was 12 footcandles, the results were an average score of 9 mm using binocular vision, and an average of score 22 mm using monocular vision, for a ratio of 1:2.45. The best depth perception score (8 mm) was using binocular vision, with a background illumination of 1 footcandle, and the worst depth perception score (22 mm) was when the subjects were using monocular vision with the background illumination at 6 or 12 footcandles. This indicates that the best average depth perception occurred when binocular vision was used, at the lowest level of background illumination, and the worst average depth perception occurred when using monocular vision at the highest levels of illumination.

Berry (1948) found in the vernier condition, monocular vision was in general equal to binocular vision. But for stereoscopic depth perception and real depth perception, binocular vision was superior to monocular vision The results obtained when comparing monocular and binocular vision in determining depth perception in the current experiment agree with those found by Berry.

The Chalmers (1952) experiment studied monocular and binocular depth cues over distance. He found that monocular vision was sufficient up to about 25 ft, but beyond that binocular vision was necessary to make accurate depth observations.

Barrett and Williamson (1966) had their subjects relate the quality or sensation of depth of a three dimensional scene. They conducted their experiment in such a way that the subjects did not know when they were viewing a scene using one or two eyes. The subjects reported the scene and the perception of depth as having better quality when the scene was viewed using binocular vision.

The results of the present experiment agree with Chalmers (1952), and Barrett and Williamson (1966), depth perception is better when binocular vision is used.

The second part of the experiment was conducted to determine the effect of background illumination on depth perception. It was expected that as background illumination was increased, depth perception would increase. Support was not found for this hypothesis.

Mueller and Lloyd's (1948) research indicated that depth perception increased as the level of light increased, until at high intensities there was no further improvement in depth perception. Berry, Riggs, and Duncan (1950) also concluded that depth perception was aided by increased levels of light.

The results of this experiment do not agree with the results of experiments performed by Mueller and Lloyd, or Berry et al. The present experiment indicated that the level of background illumination made no difference in depth perception.

Meaningful comparisons cannot be made between the present experiment and previous experiments by just comparing the numbers obtained. This is because of the different units of measure used in the experiments. Previous experimenters plotted their results millilamberts, the present experiment used footcandles. However some meaningful information may be obtained by comparing the graphs. Berry et al. and Coules noted that as the light source became brighter, to a point, depth perception improved, then as the light got even brighter, depth perception worsened. An examination of Figure 2 reveals approximately the same phenomenon of depth perception under the monocular condition. Perception of depth started off with a low score for accuracy of depth judgment, it got better as the background illumination increased, then worsened at the highest intensities of background illumination.

It was further demonstrated, that as a group, the 15 subjects who participated in the present experiment had equivalent depth perception, and that they performed no better or worse than each other.

This study of depth perception involved the use of a limited number of depth cues. In this experiment the only depth cues available were those in the Howard-Dolman box. Depth perception is more than what can be observed directly from an object of interest; it is also how that object relates to the rest of its surroundings. Depth perception is used for everything from picking up a pencil to driving an automobile. The ability to perceive depth accurately allows us to move freely in and manipulate our environment.



References

Barrett, G. V. & Williamson, T. R. (1966). Sensation of depth with one or two eyes. Perceptual and Motor Skills, 23, 895-899.
Berry, R. N. (1948). Quantitative relations among vernier, real depth and stereoscopic depth acuity. Journal of Experimental Psychology, 38, 708-721.
Berry, R. N., Riggs, L. A. & Duncan, C. P. (1950). The relation of vernier and discrimination to field brightness. Journal of Experimental Psychology, 40, 349-354.
Chalmers, E. L., Jr., (1952). Monocular and binocular cues in the perception of size and distance. American Journal of Psychology, 65, 415-423.
Coules, J. (1952). Effects of photometric brightness on judgment of distance. Journal of Experimental Psychology, 50, 19-25.
Matlin, M. W. (1988). Sensation and Perception (2nd ed.).
Boston: Allyn and Bacon, Inc.
Mueller, C. G. & Lloyd, V. V. (1948). Stereoscopic acuity for various levels of illumination. Proceedings of National Academy of Science, 34, 223-227.
Rathus, S. A. (1990). Psychology (4th ed.). Fort Worth: Holt, Rinehart and Winston.


[To Andy's Psychology Menu Page][To Welcome Page]


How you can help... I hope you enjoyed these pages and found them useful in some way. Now, here's how you can help. Some of you have expressed a desire to help support these pages financially, I thought about it, and here it is... All I'm asking is a small donation ($3.00 - $5.00). Use of these pages is still free, no charge, you may continue to use them free of charge, as much as you want, as often as you want, anytime you want... But you can also use the links below to donate and help keep these pages here, and help the site expand. You may use your PayPal account, or most any credit card. Thank you for your continued support.

Select this button to donate $3.00
Select this button to donate $5.00



The purpose of this information: This information was first published as an experiment in publishing on the WWW. These pages are case studies of a fictional person. I composed these papers as an assignment for a graduate level Theories of Personality, Counseling Psychology class. I'm not claiming that any of this material is 100% correct. I make no representation as to the usefulness of this material. I hope that these pages are of some use to you, either educational, informative, or at least entertaining. If you have any comments about these pages please contact me by e-mail. Thank you for reading.


To comment about these pages contact Andy at andyda@earthlink.net.