The perception of the size of an object is as basic a component of our perception of that object as is its shape, color, or location in three-dimension space. How does our visual system create this feature of our visual experience of an object?
One obvious cue to the size of an object is the size of the retinal image projected onto the retina when we look at the object. As you have learned already, when we look at an object, an upside down image of that object is projected by the lens onto the retina. That image itself, of course, has a size. In general, the bigger an object, the bigger its retinal image size. For example, in the picture below you can see the size of the retinal image of the plane on the retina. If the plane were twice as big, its retinal image would also be twice as big. There is no question that your visual system creates the perception of the size of an object based, in part, on retinal image size information.
There is, however, just one problem, a phenomenon know as Emmert's law. According to Emmert's law, retinal image size varies not only as a function of the size of an object, but also as a function of the distance of the object from the viewer. For any particular object, the closer the object is to the viewer, the larger the image will be of the object on the retina. For example, in the picture with the airplane, it is easy to see that if the airplane were to move further away, the image on the retina of the eye would get smaller. If the airplane were to get closer, the image would get larger. Because of Emmert's law, it is quite possible for two objects of very different size to project the same size retinal image -if the larger object is also further away (as in the picture here of two airplanes of different size).
If we were to assume that retinal image size information was the only information used by the visual system to create a perception of the size of an object, then we could make two important predictions about the way the world should appear to us.
- Large objects would often look smaller to us than small objects (in those situations in which the larger object is further away).
- Whenever an object moved closer to us, we would see the object as literally getting bigger; and whenever an object moved further from us, we would see the object as literally shrinking.
It should be obvious to you that neither of these descriptions corresponds with the way the world really looks to us. We usually see large objects as being bigger than small objects, even when the large objects are further away. We do not perceive objects as changing their size when they change their distance from us. A phenomenon called "size constancy" describes this feature of our perception of size. Size constancy is "the tendency to perceive objects as maintaining their size despite changes in their distance (that is, despite changes in their retinal image size)."
How is size constancy achieved? The only possibility is that your size perception system must utilize two cues when creating the perception of size: retinal image size information and distance information. Your visual system integrates these two kinds of information when creating the perception of the size of an object. It is as if, when an object moves further from us, your visual system "knows" that even though the retinal image is getting smaller, the reason for this is because the object is moving away from us.
One implication of the way in which size perception is created is that if the visual system is tricked into misjudging the distance of an object (or misjudging the relative distances of two objects), there will be a corresponding error made in the perception of the size of the object. One example of this phenomenon involves an interesting error in size perception created by an "Ames Room." The two pictures below were taken of two women standing in different corners of an Ames room. As you can see, the woman with the red jacket looks much bigger when she is standing on the right than when she is standing on the left. Why?
The obvious reason is that the size of her image on your retina is much larger when she is on the right than when she is on the left. Can your figure out why her image is so much larger when she is on the right than on the left? Obviously it is not because she has grown by walking across the room. The answer is that the far right hand corner of the room is actually much closer to the camera than is the far left hand corner. Because she is closer on the right than on the left, her image is much larger on the right than on the left. Of course, this is what always happens when someone gets closer to us, but we do not normally see the person growing in size the way people appear to do when they walk across an Ames room.
Why do we see someone moving from the left to the right in an Ames room as literally getting bigger? The reason is because, even though the person is getting closer to us, the room has been specially constructed to hide the fact that the far left hand corner is further from us than the far right hand corner. The room has been constructed to fool our visual system into thinking the two corners of the room are the same distance from us. If the two corners are treated by the visual system as being the same distance from us, then when someone walks from the left to the right (and their retinal image gets much bigger because the person is actually also getting closer to us) our visual system will interpret the change in retinal image size as a change in the object's actual size.
A related effect occurs in a number of classic visual illusions. All these illusions operate by tricking the visual system into judging two objects that are, in fact, the same distance from us, as if one of the objects was actually further from us than the other object. For example, in the first illusion below, the three cylinders are actually all the exact same size. They are also the exact same distance from your eyes (they are all on the flat surface of your computer screen). Why, then, does the rightmost cylinder look to be the largest of the three cylinders and the leftmost cylinder appear to be small than the other cylinders?
The reason is because the lines in the picture trick your visual system into judging that the cylinders are at different distances from you (leftmost closest, rightmost furthest away). Because the retinal images corresponding to the three cylinders are exactly the same size, if one of the cylinders is judged by our distance perception system as being furthest away, that cylinder will also be perceived to be largest.
An identical explanation can be used to explain the Ponzo (or train tracks) illusion and the Muller-Lyer illusion. In the Ponzo illusion on the left, the two horizontal lines are actually identical in length. Most people see the top line as bigger, because the diagonal lines trick your visual system into treating the top line as being further away. Similarly, in the Muller-Lyer illusion on the right, the two lines are identical in length. The arrowheads at the ends of the lines, however, trick your visual system into treating the vertical line on the right as if it were further away, and therefore you perceive the line on the right as being bigger.
The following pictures were taken on the salt flats of Boilivia. Because of the wide expanse of uniform whiteness, the area is lacking in many of the monocular depth cues that our visual system uses for locating objects in depth. As a result, it is possible to take pictures that create odd illusions of size and location.
The first picture is a picture of a cup. It looks to be of normal size, and it really is of normal size.
Now look below.
Now the cup looks big enough to hold a person. Can you figure out the secret of the illusion? The next few pictures should help.
Notice in the picture above how small the man looks who looks as if he is being eaten. That is because, in reality, he is much further away from the camera than is the man lying down, and therefore the size of his image is much smaller than that of the man lying down. Normally, when we see someone far away like this, the person does NOT look tiny. Instead, the size constancy system take distance into account, and compensates for the distance-related reduction in image size. In this picture, however, there are no good depth cues available to signal that the one person is actually much further away than the other.
In the next picture, three "little" men are balancing on the hands and head of one person of normal size.
Of course, the three "little" men are simply further away from the camera. A slightly different version of this picture is below -- but in the next picture, it is a little easier for your visual system to judge that the three standing males are further away than the kneeling male.
In this picture, if you focus on the shadows of the three standing males, you might start to see them as being further away. The shadows help locate the images closer to their real location in depth (that is, further away than the kneeling male). The shadows were electronically removed from the previous picture, and most people judge that the illusion is stronger in the previous picture. Also notice that the three standing males actually look bigger in this picture than in the previous one. That is because the three standing males are perceived as being further away in this picture than in the previous picture. As a result, the size constancy system compensates more in the second picture to create the perception that the males are larger.
The effect is also not as powerful in the next picture. Notice that the image of the man is slightly out of focus. That lack of clarity leads your visual system to tend to locate him as being further away than the pack of cigarettes.
Of course, when you are actually in the salt flats, you do not experience these kinds of illusions as much as you do with the pictures. The reason for this difference is that the pictures show the view as seen by a single un-moving eye. If you are actually there, however, you look out at each scene with two eyes (providing retinal disparity information) and you move your gaze by moving your eyes and head and body (providing motion parallax cues to the real depth relationships).