Depth Cues, Illusions and Constancies
What Are Depth Cues?
Your eyes only ever receive flat, two-dimensional images — one from each retina. Yet you experience a rich, three-dimensional world full of depth. Psychologists call the pieces of information the brain uses to judge distance and depth depth cues. Some depth cues can be picked up using only one eye — these are monocular cues. Others need both eyes working together — these are binocular cues.
Monocular depth cues work with just one eye and are properties of the scene itself — they even work in a photograph or painting. Binocular cues need two eyes working together and come from the physical set-up of the visual system, not from the picture itself.
Monocular Depth Cues
Edexcel requires five monocular cues: height in the plane, relative size, superimposition, linear perspective and texture gradient. Each works using information available to a single eye.
Height in the plane means objects positioned higher in the visual field, closer to the horizon, are perceived as further away. In a photograph of a field of sheep, the sheep near the top of the image look further away than the sheep near the bottom.
Relative size works when two objects are known to be the same actual size — the one producing the smaller image on the retina is judged to be further away. Two identical cars parked at different distances will produce very different sized images, and the smaller one is seen as more distant.
Superimposition happens when one object partly covers another. The covered object is perceived as further away, and the object doing the covering is perceived as nearer. A tree trunk overlapping a fence in the background is a simple example.
Linear perspective is the way parallel lines seem to converge towards a single vanishing point as they stretch into the distance. Lines that appear closer together signal greater distance — think of railway tracks appearing to meet on the horizon.
Texture gradient describes how a repeating surface texture appears finely detailed close up and gradually smoother and less detailed further away. On a gravel path, individual stones are visible near your feet but blur into a smooth grey surface further into the distance.
The five monocular depth cues are height in the plane, relative size, superimposition, linear perspective and texture gradient. All five can be detected using only one eye and all work in a flat photograph.
Binocular Depth Cues: Stereopsis
Because your eyes sit roughly six centimetres apart, each eye receives a very slightly different view of the same scene. The brain compares the two images and uses the difference between them to calculate depth. This is called stereopsis. The closer an object is, the bigger the difference between what the left eye sees and what the right eye sees.
You can notice stereopsis by holding one finger close to your face and looking at it with only your left eye, then only your right eye, blinking between the two. The finger appears to jump sideways — that is the disparity between the two retinal images. Stereopsis works best for close objects; beyond about six metres, the difference between the two eyes’ images becomes too small for the brain to use, which is why distant depth judgements rely more on monocular cues.
Visual Illusions
A visual illusion occurs when the brain’s normal processes for interpreting depth cues and images produce a perception that does not match physical reality. Edexcel groups illusions into three types: fictions, ambiguous figures and distortions.
Fictions occur when the brain perceives something that is not actually there. In the Kanizsa triangle, viewers see a solid white triangle even though no triangle is actually drawn — the brain fills in edges that do not physically exist.
Ambiguous figures are images that can be perceived in two different ways, with the brain flipping between interpretations. The Necker cube can be seen pointing in one of two directions, and the Rubin vase flips between a vase and two faces looking at each other.
Distortions happen when the true physical properties of an image, such as length, are misjudged because of misleading surrounding cues. In the Müller-Lyer illusion, two lines of equal length appear different lengths because of the arrowheads added to their ends.
Illusions exploit the same rules the brain normally relies on to judge depth and size. In the Müller-Lyer illusion, the arrowheads mimic the converging and diverging lines used in linear perspective, so the brain misapplies a normally useful depth cue and misjudges the length of the line. Illusions are important evidence in the wider debate over whether perception is driven mainly by the image itself or by the brain’s prior expectations.
Visual Constancies
Even though the image on your retina constantly changes as you move, view objects from different angles, or view them under different lighting, your perception of an object’s properties usually stays the same. This is called visual constancy. Edexcel requires three types: shape, colour and size.
Shape constancy is perceiving an object as keeping the same shape even though its retinal image changes as your viewing angle changes. A door is perceived as rectangular whether you view it face-on or from an angle, even though the image it casts on your retina is trapezoid-shaped when viewed from the side.
Colour constancy is perceiving an object as keeping the same colour even when the lighting changes the wavelengths of light reflected from it. A white shirt still looks white in dim evening light and in bright sunlight, even though the physical light reaching your eye is very different in each case.
Size constancy is perceiving an object as staying the same physical size even though its retinal image shrinks as it moves further away. A friend walking away from you is not perceived as shrinking — the brain uses distance information to correct the size estimate.
The three visual constancies are shape (an object's shape is perceived as stable despite a changing viewing angle), colour (an object's colour is perceived as stable despite changing light), and size (an object's size is perceived as stable despite changing distance).