Figure-Ground Perception: How the Brain Separates Objects from Backgrounds

A border by itself does not tell you which side is the object. Your visual system has to make that decision — and sometimes the exact same image can flip between two answers.

Figure-ground perception is the process that separates a visual scene into a figure — the region treated as the object — and a ground, which is experienced more like the background continuing behind it.

That sounds simple until the image is ambiguous. If two regions share the same border, which one owns the edge? Which one has the shape? Which one appears in front?

Try the figure below. Do you see a central vase first, or two profiles facing each other? The pixels do not change. What changes is the assignment of figure and ground.

Reversible images are useful because they expose a job that is normally invisible. In everyday vision, the brain usually settles on one organization so quickly that you never notice an alternative was possible.

What Does “Figure” Actually Mean?

The figure is more than the part you happen to stare at. In classic figure-ground organization, the figure tends to appear shaped by the shared boundary and positioned in front, while the ground appears to continue behind it. Researchers describe this in terms of border ownership: the shared edge is perceptually assigned to one side.

A major review by Wagemans and colleagues describes figure-ground organization as distinct from perceptual grouping. Grouping determines which elements go together; figure-ground organization determines how those organized regions are interpreted as surfaces and shapes in a scene.

That distinction is why this topic sits next to, but is not the same as, visual grouping and the Gestalt principles of perception.

What Makes One Side Become the Figure?

There is no single cue that always decides the winner. Several image properties can bias one region toward figure status. A smaller surrounded region often looks like the object against a larger surrounding field. Convex regions tend to be favored over concave alternatives. Symmetry, contrast, depth information, and the location of a region in the display can also matter.

Past experience can contribute too. In experiments by Mary Peterson and Bradley Gibson, object-recognition processes influenced figure-ground organization for displays defined by several types of contours. Their work challenged a strictly bottom-up sequence in which the visual system would first assign figure and ground and only afterward begin recognizing shapes.

In other words, the question “what does this edge probably belong to?” can be influenced by what the adjoining regions resemble.

Border Ownership Is Represented Surprisingly Early

Figure-ground perception is not merely a conscious interpretation added after the visual system has finished processing an image. Neural recordings show signals related to border ownership relatively early in visual cortex.

In a study of behaving macaques viewing complex natural scenes, Jonathan Williford and Rüdiger von der Heydt found that about half of the sampled V2 neurons were selective for which side of a contour owned the border in natural images. The border-ownership signal emerged at roughly 70 milliseconds after stimulus onset — only about 30 milliseconds after the neurons began responding.

That is striking because natural scenes are messy. The edge of a real object may run through shadows, texture changes, other surfaces, and partial occlusion. Yet the study found context-sensitive border assignment appearing very quickly.

Attention Favors the Figure

Once the visual system has organized a region as a figure, that organization can help structure attention.

Qiu, Sugihara, and von der Heydt recorded from neurons in macaque area V2 and found that figure-ground organization and top-down attention converged in the same cells. Attention produced stronger modulation when the attended figure lay on a neuron's preferred side of border ownership.

The important point for everyday perception is not that every figure automatically captures all attention. It is that the visual system's organization of the scene provides structure that attention can use. “Object” and “background” are different starting points for selection.

If you want to test that neighboring ability directly, CT's live Attention Test focuses on selecting relevant information while resisting distraction.

Why Reversible Figures Keep Flipping

In an ambiguous display, neither interpretation has enough evidence to win permanently. The shared contours support two plausible organizations, but you generally experience one at a time. After a few seconds, perception may reverse.

The classic face-vase style image works because the same central boundary can be assigned to the central region or to the two outer regions. When ownership switches, so does the apparent object. A vase becomes background while the profiles become figures, or the profiles recede while the vase takes shape.

This is a visual illusion, but it reveals an ordinary perceptual requirement: every crowded scene contains borders, and the visual system constantly has to decide which surfaces those borders belong to.

Figure-Ground Perception in Real Scenes

Think about a person standing in front of a wall. The outline of the person's shoulder is also an edge in the wall region of the retinal image. Perception has to assign that edge to the person so that the wall can be experienced as continuing behind them.

The same problem appears when tree branches overlap buildings, a cup blocks part of a plate, or text sits on a patterned background. Strong figure-ground separation makes the intended object easy to select. Weak separation creates camouflage and visual clutter.

This is one reason designers care about contrast and boundaries, and why camouflage works by disrupting reliable figure-ground cues. The difficulty is not simply “can the eye detect an edge?” but “can the visual system decide what that edge belongs to?”

Figure-Ground Perception and Optical Illusions

Reversible figures are only one way perception can reveal its assumptions. Other illusions manipulate size, context, depth, contour completion, motion, or contrast. The broader Optical Illusions section collects CT's visual-perception demonstrations and explanations.

You can also try the live Optical Illusion Test, which turns several classic perceptual distortions into direct judgments rather than simply showing you pictures.

Why It Matters for Pattern Recognition

Pattern recognition depends on having useful units to compare. Figure-ground organization helps decide which region is an object worth tracking and which information belongs to the background. A shape cannot rotate, repeat, or transform in your mental model until perception has established what counts as the shape.

That is why figure-ground perception belongs naturally inside the wider Pattern Recognition section. It is one of the perceptual steps that makes higher-level rule detection possible.

For more examples of how separate visual elements become organized, continue with Gestalt Principles of Perception. For broader assessment-style tasks, the brain tests collection covers attention, memory, reasoning, perception, and processing speed, while Cognitive Train's homepage brings together the wider set of brain training and cognitive training tools.