Lesson 6 of 8 · 18 min
Gestalt principles and depth cues
Explain how figure-ground, proximity, similarity, closure and continuity group parts into wholes, and how binocular and monocular depth cues let you perceive a three-dimensional world.
What you will be able to do
- Explain and identify figure-ground, proximity, similarity, closure and continuity.
- Distinguish binocular depth cues (retinal disparity, convergence) from monocular cues.
- Identify the pictorial depth cues in a scene: linear perspective, interposition, relative size, texture gradient and height in the visual field.
The retina records a flat mosaic of light: no objects, no edges that belong to anything, no depth. Two sets of visual perception principles turn that mosaic into a world of separate things at different distances. Gestalt principles group the parts into wholes, and depth cues place the wholes in three dimensions.
Gestalt principles
The Gestalt psychologists, working in Germany early in the twentieth century, argued that "the whole is different from the sum of its parts": we perceive organised wholes, not collections of dots and lines. They described the rules the brain uses to organise them.
Proximity, similarity, closure and continuity.
| Principle | We tend to perceive… | Everyday example |
|---|---|---|
| Figure-ground | …a scene as a figure (the object we attend to) standing out against a background. The same edge belongs to the figure, not the ground. | Words on this page are figure; the page is ground. In the vase-and-faces drawing the two readings swap. |
| Proximity | …elements that are close together as belonging to one group. | Dots spaced in pairs of columns are seen as double columns; words are groups of letters separated by spaces. |
| Similarity | …elements that are alike in colour, shape or size as a group. | Players in the same jersey are seen as one team. |
| Closure | …incomplete figures as complete, filling in the gaps. | A circle with gaps is still seen as a circle; a logo made of separate strokes is read as one shape. |
| Continuity | …lines and patterns as continuing smoothly in the same direction, rather than changing abruptly. | Two crossing lines are seen as two lines, not four segments meeting at a point; a rope coiled behind a post is seen as one rope. |
Change the spacing until rows become columns. Then try similarity against proximity, closure, continuity and figure-ground with the picker.
At a netball match, a spectator instantly sees the players as two teams, even when they are mixed up across the court. Which Gestalt principle best explains this?
Worked example
Identify Gestalt principles in a logo
A school logo is made of twelve separate short dashes arranged in a ring with small gaps between them. Inside the ring, three dots on the left are coloured blue and three dots on the right are coloured gold, all evenly spaced in one row.
Identify the Gestalt principle that explains how each part is perceived: (a) the ring, (b) the dots.
Step 1: The ring
The dashes have gaps, but we perceive one complete ring: the brain fills in the missing parts. This is closure.
Step 2: The dots
The dots are evenly spaced, so proximity does not split them. We perceive two groups of three, blue and gold, because elements alike in colour are grouped: similarity.
When spacing is even, any grouping you see must come from something other than proximity.
Try it yourself first, then check the working one step at a time.
Answer
(a) Closure. (b) Similarity (of colour).
Method
Ask what the parts have in common: nearness (proximity), likeness (similarity), a gap filled (closure), a smooth path (continuity), or standing out from a background (figure-ground).
Watch out for
Answering proximity for the dots: they are evenly spaced.
Depth cues
Your retina is curved but effectively flat: it records an image in two dimensions. Yet you see a world in three. The brain judges distance from depth cues, information from the eyes and the scene. Some need two eyes (binocular); the rest work with one eye (monocular), and most of these work in flat pictures too, so they are also called pictorial cues.
| Cue | How it works | Most useful for |
|---|---|---|
| Retinal disparity | Your eyes are about 6 cm apart, so each retina gets a slightly different image. The brain compares them: the bigger the difference (disparity), the closer the object. | Objects within a few metres, such as threading a needle or catching a ball. |
| Convergence | To look at a near object the eyes turn inwards. The brain uses the tension in the eye muscles: more turning means a closer object. | Close objects, within a few metres. |
Close one eye and the world does not go flat, because of the monocular cues. Accommodation, the change in the lens's shape to focus at different distances, is a monocular cue from the eye itself, useful only for objects very close by. The others are in the scene.
Pictorial depth cues in one scene: linear perspective (A), texture gradient (B), relative size and height in the visual field (C), interposition (D).
| Cue | What you see | How the brain reads it |
|---|---|---|
| Linear perspective | Parallel lines, such as road edges or railway tracks, appear to converge towards the horizon. | The closer together the lines, the further away. |
| Texture gradient | Surfaces look coarse and detailed close up and finer and denser further away. | Finer, more tightly packed texture means further away. |
| Relative size | Of two objects we know to be about the same size, one makes a smaller image. | The one with the smaller image is further away. |
| Interposition (overlap) | One object partly hides another. | The object that is covered is further away. |
| Height in the visual field | Of objects below the horizon, some sit higher in the visual field, closer to the horizon. | Higher, closer to the horizon, means further away. |
Turn the five pictorial cues off one by one and watch the scene flatten. Then move the object nearer in the binocular panel and watch convergence and disparity grow.
Classify each depth cue as binocular or monocular.
In a photograph of a jetty, the planks near the camera look wide with clear gaps between them, and further out they blur into a fine, even surface. Which depth cue is this?
Worked example
Explain depth perception with one eye
After losing the sight of one eye in an accident, a man can still drive safely on the highway, but at first he has trouble pouring water into a glass at arm's length.
Explain both observations with reference to depth cues. (4 marks)
Step 1: Which cues he has lost
With one eye he has lost the binocular cues: retinal disparity (comparing the two retinal images) and convergence (the eyes turning in).
Step 2: Why near tasks suffer
Binocular cues are most useful for near objects, within a few metres, so judging the distance of the glass at arm's length becomes harder.
Step 3: Why driving is still possible
Highway driving depends on distances far beyond the range of binocular cues. Monocular pictorial cues such as linear perspective (converging road edges), relative size and height in the visual field (other cars), texture gradient and interposition work as well with one eye.
Every cue that works in a photograph works for one eye.
Step 4: Add accommodation
He still has accommodation, a monocular cue for close objects, and with practice he learns to rely on it and on pictorial cues for near tasks too.
Try it yourself first, then check the working one step at a time.
Answer
He lost retinal disparity and convergence, binocular cues that matter most for near objects, so pouring at arm's length is hard. Driving relies on far distances judged by monocular cues (linear perspective, relative size, height in the visual field, texture gradient, interposition), which work with one eye.
Method
Identify the cues lost and the cues kept.
Link each to the distances it works for.
Watch out for
Saying a person with one eye has no depth perception.
Common mistake
Pictorial cues are monocular
Every cue that works in a painting or a photograph works for one eye, so it is monocular. Only retinal disparity and convergence need two eyes. Accommodation is monocular too, but it is not a pictorial cue: a picture's surface is all at one distance.
Summary
Gestalt principles organise parts into wholes: figure-ground, proximity, similarity, closure and continuity.
Depth cues let a flat retinal image be perceived in three dimensions.
Binocular cues need both eyes: retinal disparity (bigger difference between the two images, closer object) and convergence (more inward turning, closer object).
Monocular cues work with one eye: accommodation, and the pictorial cues linear perspective, texture gradient, relative size, interposition and height in the visual field.
Binocular cues work best for near objects; pictorial cues work at any distance and in pictures.
Finished the lesson?
Try the 3 quick checks above and the lesson completes itself, or mark it complete now.