A visual illusion may refers to the case where what we perceive differs from what we suppose to be correct. As Nicholas et al. (2001) note that the idea of a visual illusion presupposes that the object or pattern concerned would be different under other conditions. One view of illusions is that they can be used as tools to probe the mechanisms of visual perception, because perceptual errors give us clues about the way in which normal perception takes place .
Gregory (1997) explains that the act of perceiving is a dynamic process. He goes on arguing that that perception is the brain's search for the best interpretation of the data that is being presented. Sometimes the perceptual hypothesis is incorrect and an illusion occurs. Most of the theories for visual illusion are categorized in two groups. The first group attributes illusions to innate and fixed physiological neural cell activities. These activities involve lower level and more independent functions of human. These functions are collectively referred to as early vision or low level vision. Theories on the second group are concerned with higher levels such as the feelings, will or intelligent judgment of humans.
The high and low levels of processing can be explained as top-down and bottom-up process respectively. When a stimuli is presented to the participant, it causes certain internal cognitive process to occur. This process finally produces the required response or answer. As explained buy M. Eysenk (2001), processing directly affected by the stimulus input is usually described as bottom-up processing. On the other hand, top-down processing is the one that is influenced by the individual's expectations and knowledge rather than the stimulus itself.
Contructivist theorists emphasize the role of top-down process in visual perception, whereas direct theorists emphasize bottom-up processes and the richness of the information contained. We will refer lower level processing as "bottom-down" knowledge and high level as "top-down" processing.
As Shelly Wu stresses, nothing can illustrate the role of higher process (top-down) in the human visual system better than illusions. Ambiguous figures are extremely important for showing the dynamics of perception, the searching for hypotheses of objects that might or might mot be in the external world (M. Eyesenk 2001). Moreover, they allow us to separate effects of bottom up signals from the eyes, from top down knowledge assumptions. Gregory (1998) states that the more the top down contribution, the less direct is perception.
The hollow mask is a good example of top-down process, where top-down object knowledge can dominate bottom-up signals
Hollow Face Illusion
The top left image corresponds to viewing the outside of the mask (e.g., nose pointing at you). The image rotates 180 degrees (bottom right image), so that you are viewing the inside of the mask (e.g., nose pointing away from you, as if you were about to put the mask on). When viewing the inside of a mask or shape of a human face, the face is frequently perceived as being a normal (convex) face, instead of the veridical, hollow (concave) face. This is because the mask is hollow but it is hard to convince your eyes to see it in this way. Thus, familiarity with the shape of faces dominates perception, even when in conflict with stereo depth cues.
The strong visual bias of favoring seeing a hollow mask as a normal convex face is evidence for the power of top-down knowledge for vision (Gregory 1970). Gregory (1997) in his work on "Knowledge and Perception in visual illusions" stresses out that there is a weaker general tendency for any object to be seen as convex, probably because most objects are convex. The effect is weaker when the mask is placed upside down, strongest for a typical face.
Ames Room illusion
There are two illusions associated with the Ames Room. First the room appears cubic when viewed monocularly from a special viewing point (the true shape of the room is trapezoidal). Secondly, within an Ames Room people or objects can appear to grow or shrink when moving from one corner to the other.
Research indicates that the Ames room illusion can be explained by the lack of cues normally used in three-dimensional shape constancy (Dorward & Day, 1997). When you look through a peephole into an Ames Room, the room looks normal and cubic, but its true shape is cleverly distorted. The floor, ceiling, some walls, and the far windows are actually trapezoidal surfaces. Although the floor appears level, it is actually at an incline (the far left corner is much lower than the near right corner). The walls appear perpendicular to the floor, although they are actually slanted outwards.
The right diagram shows how the Ames Room forms an identical image of a normal cubic room on your retina. If a straight line (representing a ray of light) is drawn from one corner of an imaginary cubic room to your eye, the corner can meet this ray at any point along its length and still appear cubic. Since the two visible corners of the room subtend the same visual angle to the eye through the peephole, the two corners appear to be the same size and distance away. The left corner, however, is actually twice as far away as the right corner. When the view sees the room from another angle the true shape of the room is revealed.
The retinal image produced by the distorted room is identical with that of a normal cubic room. Seckel Al states that your visual system relies partly on past experience with normal cubic rooms to judge the shape of the room. Therefore, we conclude that this illusion is explained by of top down processing, i.e., that your visual system resolves ambiguity based upon knowledge of the external world.
The famous Müller-Lyer illusion shows us how previous knowledge derived from the perception of three dimensional objects is applied inappropriately to the perception of 2 dimensional objects (Gregory 1970). In this illusion, our perception does not match physical reality since we perceive the left vertical line longer than the left one, when in fact are of the same size.
The traditional perspective theory states that these figures suggest depth, and if this suggestion is followed up, the most distant features appear larger. One Gregory's explanation is that is that the figures are treated as three-dimensional objects, although they seem flat (Gregory 1970). The left figure can be seen as the inside corners of a room, where right figure is seen as the outside corners of a building, therefore, it seems further away . Because of the size of the retinal image is the same for both vertical lines, the principle of size constancy tells us that the line that is further away must be longer.
However, some theorist argued that Gregory's theory is incomplete. The Müller-Lyer illusion remains when the fins on the figures are replaced with other attachments (e.g. circles). Matlin and Foley (1997) introducing the incorrect comparison theory state that "our perception of visual illusions is influenced by parts of the figure not being judged". In the illusion, lines may seem shorter/larger because they form a part of a larger or smaller object.
This theory, unlike the one introduced by Gregory, assumes that direct perception does not require an individual to be active in processing information (top-down processing) in order to construct a perception from cues. Instead, all the information for perception lies in the environment. The person's perception is immediate and spontaneous.
Therefore, we can argue that the Müller-Lyer illusion uses both high and low level of processing. When the lines are perceived as corners of a building, high level(top down) of processing is used. On the other hand, when the lines are perceived as part of a smaller/larger object, low level of processing (bottom-up) is used.
The next two illusions will help us understand the low level processing in visual perception. In the illusion of L.Herman, hence the name Hermann grin illusion, we see dark spots in the intersections of white horizontal and vertical stripes. When we try to look directly at the spot, we cannot see it.
Hermann grin illusion
The spots are illusory, they do not exist. The Hermann Grid illusion occurs because of interactions between the grid of squares and a process in your visual system called lateral inhibition. It is believed that these spots are products of centre/surround antagonism within receptive fields of retinal ganglion cells. Looking at the lower part of the diagram, we can compare how the retinal image of the grid would affect the two receptive fields shown, one being stimulated by an intersection and the other being stimulated by part of a white stripe that is not at an intersection.
As Sekuler et al. (1994) explain, the receptive field that lies at the intersection of the white cross has more light falling on its inhibitory surround than does the receptive field that lies between the two black squares. Consequently, the excitatory center of this receptive field between the squares yields a stronger response than that which lies at the intersection of the white cross.
The upper right of the diagram shows that receptive fields in the central fovea are much smaller than in the rest of the retina. That is why we cannot see the dark spot when we look directly at the intersection of the white cross.
Adelson's journal "Lightness Perception and Lightness Illusion" concludes that illusions of lightness and brightness can help reveal the nature of lightness computation in the human visual system. It appears that low-level, mid-level, and high-level factors can all be involved.
We now see another illusion on lightness perception, the "Mach band illusion". As explained by Adelson, when a spatial ramp in luminance suddenly changes slope, an illusory light or dark band appears. The illusion was illustrated by the famous artist Vasarely (left figure (a)).
The image consists of a set of nested squares. Each square is a constant luminance. The pattern gives the illusion of a glowing X along the diagonals, even though the corners of the squares are no brighter than the straight parts. When a center-surround filter is run over this pattern, it produces the image on the left (b). The filter output makes the bright diagonals explicit.
David Marr, a British psychologist who made important contributions to the study of visual processing, once said that perception is the construction of a description. Since we have seen that illusions directly influence our descriptions of objects and phenomena, we can conclude that illusions can be really useful to our understanding. Illusions can help us to understand perception because they offer excellent clues that tell when and how our normal perception fails.
Visual illusions should not be considered as phenomena that point to the inadequacies of our visual and perceptual systems. We should think of an illusion as a phenomenon that allows us to become consciously aware of the intricate process that is always unfolding behind the scenes .
1. Gregory, R. L. (1998). Eye and Brain (fifth edition). Oxford University Press
2. Sekuler, R. and Blake, R. (1994). Perception (Third Edition). New York: McGraw Hill
3. Hill H & Bruce V (1994) "A comparison between the hollow-face and hollow-potato' illusions", Perception 23, 1335-1337
4. Ames, A. (1952) The Ames Demonstrations in Perception, New York, Hafner Publishing
5. Richard L Gregory (1997) Knowledge in perception and illusion, Department of Psychology, University of Bristol, 8 Woodland Road, Bristol
6. Matlin, M. W., & Foley, H. J. (1997). Sensation and Perception (4th edition). Boston: Allyn & Bacon
7. http://psylux.psych.tudresden.de/i1/kaw/diverses%20Material/www.illusionworks.com/html/ames_room.html 21 Nov. 2002
8. Eysenck, W. Michael, (2001). Principles of Cognitive Psychology. East Sussex: Psychology Press
9. Edward H. Adelson, (2000). Lightness Perception and Lightness Illusions. Cambridge, MA: MIT Press, 339-351, 2000
10. http://www.kyb.tuebingen.mpg.de/publication.html?publ=80 20 Nov. 2002
11. http://psychology.about.com/msubindex_perception.htm 25 Nov. 2002
12. http://www.lems.brown.edu/vision/people/leymarie/SkiP/AmesRoom.html 25 Nov. 2002