Depth Perception

Visual perception is the process through which the human visual system interprets light and retinal images in order to comprehend objects, scenes, images and spatial reality. It enables us to identify apparent size, shape, position, orientation, colour, distance and depth, and to organise changing visual information into a meaningful and relatively stable visual world.

Vision is therefore much more than the optical formation of an image upon the retina. The retinal image provides essential visual information, but this information must be processed and interpreted through physiological and psychological mechanisms. Human visual perception combines optical image formation with visual processing, knowledge, context, memory, depth cues, movement and other perceptual factors.

Within the Perspective Research Centre framework, visual perception is especially associated with Visual Perspective Type 2: the human visual-system subclass of Visual Perspective concerned with retinal and perceptual appearances shaped by viewpoint, monocular and binocular vision, apparent size, shape, position, depth cues and visual acuity.

Visual Perception and Visual Perspective

The PRC distinguishes two related meanings of Visual Perspective:

Visual Perspective Type 1: the broad category encompassing direct and represented visual appearances of spatial objects, scenes, views and images.
Visual Perspective Type 2: the human visual-system subclass, including retinal and perceptual appearances shaped by viewpoint, monocular and binocular vision, apparent size, shape, position, depth cues and acuity.

Visual perception belongs particularly to Type 2 because it concerns how the human observer receives, organises and interprets visual information.

This distinction is important because the geometry of an image and the human perception of that image are related but not identical. A geometrically defined perspective image is subsequently viewed through a biological and perceptual system capable of interpreting, correcting, completing and sometimes misinterpreting the information it receives.

From Spatial Reality to Visual Perception

A simplified visual process begins with a spatial object or scene. Light emitted or reflected from points within that scene travels towards the eye. The cornea and lens refract and organise the incoming light so that an optical image is formed upon the retina.

Retinal photoreceptors respond to the light pattern, and visual information is transmitted through the nervous system for further processing. The resulting perception is not merely a duplicate of the retinal image. It is an interpreted visual experience of spatial Form, colour, position, motion, depth and surrounding environment.

The broad sequence can therefore be understood as:

spatial reality;
light and environmental optics;
optical image formation in the eye;
retinal response;
physiological visual processing;
psychological and perceptual interpretation; and
the experienced visual world.

Visual Perception Is Not Simply the Retinal Image

The retinal image is an optical image formed upon the light-sensitive surface of the eye. Visual perception is the interpretation and organisation of the information derived from that and related visual processes.

The distinction is fundamental. The visual world does not ordinarily appear inverted simply because the retinal image is optically inverted. Nor does the visual world appear constantly unstable merely because the eyes repeatedly move and the retinal image consequently changes.

Visual perception instead produces a comparatively coherent and stable experience from continuously changing optical information.

The Correspondence or Equivalence Problem

A central problem of visual perception is that the retinal image does not contain a simple one-to-one description of the three-dimensional spatial reality that produced it.

A light ray arriving at a particular retinal direction might theoretically have originated from an object point located at any of many distances along the corresponding spatial direction. The retinal response therefore does not by itself specify the absolute depth of the originating point.

Likewise, many different three-dimensional Forms can potentially produce similar two-dimensional projected structures from a particular viewpoint.

The visual system must consequently employ additional information, including perspective phenomena, contextual knowledge and depth cues, in order to infer the likely size, shape, distance and position of objects in spatial reality.

Spatial Perception

Spatial Perception is the human ability to perceive and interpret depth, distance and the third spatial dimension.

The Dictionary identifies three primary spatial factors that must initially be recovered or interpreted from a perspective view:

distance — the depth location of an object;
height — its location and extent in the vertical dimension; and
lateral location — its position across the visual field.

In practice, many further factors are involved, including object size, shape, orientation, angle, movement, occlusion and surrounding context.

Recovering these properties from a single view may be difficult or sometimes impossible without additional information. This is another expression of the correspondence or equivalence problem.

We Do Not Directly See Depth

One of the most important principles identified in Volume 1 is that the retina does not directly measure absolute distance.

Nevertheless, human beings experience a strongly three-dimensional visual world. The visual system achieves this by interpreting combinations of depth cues — the various optical, geometrical, physiological, environmental and perceptual forms of information that indicate distance and spatial organisation.

Volume 1 refers to these collectively as Generators of the Third Dimension.

No single depth cue must always be present. Several cues can reinforce one another, while in other circumstances one or two strong cues may be sufficient to produce a convincing impression of three-dimensional space.

The Two Principal Groups of Depth Cues

Depth perception is conventionally divided into two major groups:

Monocular depth cues: information available through one eye and therefore capable of operating in ordinary pictures, photographs and other single-view images.
Binocular or stereoscopic depth cues: information derived from the relationship between the separate views received by the two eyes.

Additional oculomotor information, particularly accommodation and vergence, can also contribute to judgements of distance.

The importance of monocular cues explains why a two-dimensional photograph, painting, film frame or linear-perspective drawing can create a compelling impression of three-dimensional space without providing separate stereoscopic images to the two eyes.

Monocular Visual Perception

Monocular perception uses information available through a single eye. It is fundamental to the relationship between perspective and depth perception.

Volume 1 identifies important monocular cues including:

Perspective of Form;
diminution of size;
aspect and foreshortening;
motion parallax;
colour;
distance fog or atmospheric perspective;
focus;
occlusion and overlapping contours;
peripheral vision;
surface texture;
horizons;
height in the visual field;
light and shade;
optic-flow patterns;
environmental invariants; and
contextual and ecological information.

These cues are particularly important for the perception of medium- and long-distance space, where stereoscopic differences become less significant.

Perspective of Form as a Depth Cue

One of the most important monocular depth cues is the Perspective of Form: the systematic change in apparent size, shape, position and orientation produced when spatial Forms are viewed from different distances and directions.

Characteristic perspective phenomena include:

diminution of apparent size with increasing distance;
Aspect Foreshortening;
Perspectival or Optical Foreshortening;
convergence of parallel directions;
formation of vanishing points;
changes of apparent shape;
diminution and degradation of Form; and
changes in the relative position of objects within the visual field.

These transformations allow the visual system to infer spatial organisation from the projected appearance of Forms.

Diminution of Size

Diminution of size is one of the most characteristic perspective depth cues.

Under otherwise comparable conditions, an object occupies a smaller visual angle as its distance from the observer increases. Repeated objects of known or assumed similar physical size can therefore establish a strong gradient of recession.

A row of equally sized figures, posts, windows, paving stones or other repeated Forms becoming progressively smaller can provide powerful information about spatial depth even in a completely flat image.

Foreshortening and Aspect

Foreshortening also contributes strongly to visual interpretation of Form.

A spatial dimension directed increasingly towards or away from the observer appears contracted relative to the same dimension viewed more nearly broadside. This orientation-dependent component is Aspect Foreshortening.

In finite-viewpoint visual perspective, dimensions extending significantly through depth are also affected by changing apparent scale with distance. This provides the additional Perspectival or Optical Foreshortening component.

The visual system can use the resulting transformed shape as evidence of the orientation and depth of the underlying spatial Form.

Occlusion and Superposition

Occlusion, also described in relevant contexts as superposition or interposition, is a powerful monocular cue.

When one object partially blocks another, the blocking object is normally interpreted as being nearer, while the partially hidden object is interpreted as lying behind it.

Occlusion establishes an ordering of depth but does not by itself specify the precise numerical distance between the objects.

Occluding contours and visible edges consequently play an important role in segmenting a scene into separate objects and surfaces.

Texture Gradients

Surface texture can provide information about depth when repeated or distributed elements undergo systematic changes across the visual field.

A textured surface extending away from the observer can show progressively smaller, denser or less distinct texture elements with increasing distance.

Such texture gradients can provide information about surface orientation, recession and spatial layout even when the precise physical size of individual texture elements is unknown.

Height in the Visual Field

The vertical position of an object within a view can also contribute to depth interpretation.

For objects resting on a common horizontal ground plane, increasing distance often corresponds to a systematic movement of their ground-contact positions towards the geometrical horizon line.

Height in the visual field can therefore contribute to the perceived organisation of a ground plane when combined with other information concerning object size, horizon position and surface geometry.

Atmospheric and Colour Cues

Distance can also change the optical appearance of Forms through the intervening atmosphere.

With increasing distance, objects may exhibit reduced contrast, diminished colour saturation, reduced clarity and loss of fine outline information. These changes form part of Atmospheric or Aerial Perspective.

Structured gradients of lost colour, contrast, visual acuity and Form can therefore function as additional depth cues.

Light and Shade

The distribution of light and shade contributes to the perception of three-dimensional Form.

Changes in illumination across a surface can help indicate whether a Form is flat, convex, concave, rounded, folded or otherwise spatially structured.

Cast shadows can additionally provide information about relationships between objects, surfaces, light sources and intervening space.

Light-and-shade information therefore operates alongside geometrical perspective rather than replacing it.

Motion Parallax

Motion Parallax is a monocular depth cue produced when the observer or surrounding objects move.

As the observer changes position, nearer objects generally undergo a larger or faster apparent positional shift than more distant objects.

The resulting pattern of relative movement provides information about depth and distance. Unlike a static perspective image, motion parallax uses changing viewpoints through time.

This makes movement itself an important component of visual perception and explains why active observation can reveal spatial information that remains ambiguous in a single stationary view.

Optic Flow

Movement of the observer through an environment produces systematic changes across the visual field, often described as optic flow.

As viewpoint changes, the positions, apparent sizes and directions of scene elements change in structured ways. These changes can provide information concerning movement, heading, surface layout and relative distance.

Visual perception is therefore not based solely upon a succession of isolated static pictures. The changing optical structure produced by observer movement is itself informative.

Environmental Invariants

When the observer moves, many visual relationships change, but others remain comparatively stable.

These stable relationships or invariants can help the visual system identify enduring spatial structures across changing views.

The contrast between changing image features and invariant environmental relationships is important to ecological theories of visual perception and to the perception of a stable spatial world during movement.

Affordances and Environmental Perception

The Dictionary also includes the ecological concept of affordances: directly perceivable relationships between an observer and the possible actions offered by objects or surfaces within an environment.

A horizontal surface may afford standing or sitting; an opening may afford passage; a handle may afford grasping.

This emphasises that visual perception is not merely the passive recognition of image geometry. It also helps organisms orient themselves and act within spatial reality.

Focus and Accommodation

The visual system can also obtain information from the optical effort required to focus upon objects at different distances.

Accommodation involves changes in the optical power of the eye as it focuses upon nearer or farther objects. The crystalline lens changes form through the action of the ciliary mechanism.

Accommodation can therefore contribute to distance perception, particularly at nearer ranges, although it operates together with numerous other visual cues.

Binocular Visual Perception

Binocular vision uses the information received by two eyes positioned at slightly different locations.

Each eye consequently observes nearby spatial Forms from a slightly different viewpoint. The visual system can compare and combine these separate images to obtain additional information about spatial relief and relative depth.

The principal binocular processes discussed in Volume 1 and the Dictionary include:

binocular disparity or binocular parallax;
stereopsis;
binocular fusion;
vergence or convergence; and
stereoscopic range information.

These are additional depth mechanisms rather than replacements for monocular perspective cues.

Binocular Disparity and Stereopsis

Because the eyes occupy different physical positions, each receives a slightly different perspective image of nearby spatial reality.

This difference is known as binocular disparity or, in related terminology, binocular parallax. The human visual system can compare the dissimilar images and use their differences as a source of relative-depth information.

The resulting perception of spatial relief is known as stereopsis.

Stereoscopic information is especially useful at close range because the angular differences between the two eye positions are then comparatively large. Its effectiveness diminishes progressively as viewing distance increases.

Binocular Fusion

The two eyes do not normally produce an experience of two completely separate worlds.

Binocular fusion describes the processes through which the visual system combines corresponding information from the two eyes into a coherent perception.

The Dictionary distinguishes sensory fusion, involving the combination of the retinal inputs, and motor fusion, involving eye movements that help maintain appropriate alignment.

Vergence and Convergence

When both eyes fixate a nearby object, they rotate towards one another. The amount of this vergence changes systematically with viewing distance.

The visual system can therefore use aspects of the eye-position relationship as another source of distance information.

Vergence should be distinguished from stereopsis. Stereopsis depends principally upon differences between the images of the two eyes, whereas vergence concerns the angular orientation of the eyes themselves.

Binocular Vision Is Not the Only Source of 3-D

A common misconception is that stereoscopic vision alone produces the perception of three-dimensional space.

This cannot be correct because paintings, drawings, photographs, television and conventional cinema images can all communicate strong impressions of spatial depth while providing essentially the same image to both eyes.

Likewise, closing one eye does not ordinarily cause the physical world to become visually flat. The observer continues to perceive extensive depth through monocular perspective, motion, occlusion, texture, known Form and other cues.

Binocular stereopsis provides valuable additional information, especially for nearby Forms and precise close-range judgements, but it is one part of a much larger system of visual depth perception.

Visual Perception at Different Distances

The relative importance of different visual cues changes with viewing distance.

At close range, binocular disparity, vergence, accommodation, motion and detailed shape information can all contribute strongly.

At greater distances, the angular differences between the two eyes become increasingly small. Monocular perspective phenomena — including relative size, foreshortening, convergence, occlusion, atmospheric change and gradients of Form — consequently become particularly important for comprehending larger-scale spatial environments.

Visual Field

The visual field is the bounded field of colours, Forms and directions available to an observer at a particular fixation and orientation.

It has a centre and peripheral limits and changes when the eye, head or body changes orientation.

The visual field should not be imagined as a uniformly detailed image. Visual acuity is greatest around the point of fixation and falls progressively towards the peripheral regions.

Human perception nevertheless combines information across this uneven field into an apparently coherent visual environment.

Central and Peripheral Vision

The central visual region provides the greatest acuity and is used for examining fine spatial detail.

The broader peripheral field has substantially lower spatial resolution but remains important for detecting movement, broad Forms, changes of illumination, orientation and events occurring away from the point of fixation.

Consequently, the visual system does not perceive every part of the surrounding world with equal clarity at one instant.

Eye movements repeatedly redirect the region of greatest acuity towards different parts of the environment, allowing visual information to be gathered sequentially.

Visual Acuity and Perception

Visual acuity concerns the ability to distinguish fine spatial detail.

The retinal image is sharpest at and near the point of fixation. Away from the foveal region, acuity falls and fine structural information becomes less distinct.

This variation is relevant to perspective because loss of detail with eccentricity or distance can itself affect the perceived Form and spatial organisation of an object or scene.

Visual perception therefore involves a trade-off between fine central detail and a much broader field of spatial awareness.

The Moving Eye and the Moving Observer

Natural visual perception is rarely based upon one permanently fixed viewpoint.

The eyes rotate, the head turns and the observer moves through space. Each change produces a new optical and retinal relationship to the surrounding environment.

Volume 1 distinguishes the field visible to a stationary eye from the much larger total spatial region explored as the eye moves. Head and body movement extend this process still further.

Human perception consequently combines information from multiple successive views into a broader understanding of the surrounding spatial world.

Visual Field and Visual World

The Dictionary records J. J. Gibson’s useful distinction between the visual field and the visual world.

The visual field is the temporary, bounded and comparatively picture-like appearance available at a particular fixation and orientation.

The visual world is the stable, extended and three-dimensional environment perceived around the observer. It does not ordinarily appear to jump or rotate every time the eyes move.

The two concepts should therefore not be treated as synonyms. The visual field changes continuously with viewing direction; the visual world is the perceptually organised environment constructed or maintained across those changing views.

Visual Space

Visual Space is the spatial organisation experienced through vision, including apparent direction, position, distance, depth, size, shape and orientation.

Visual space is related to the retinal image and visual field but is not identical to either. The retinal image is an optical image; visual space is the perceived spatial organisation derived from visual processing.

Nor should visual space automatically be equated with measurable physical object space. Optical projection and perceptual interpretation can transform apparent size, shape, position and orientation.

Physical World, Optical World and Visual World

The Dictionary distinguishes several stages or forms of spatial reality that are useful when analysing visual perception:

Physical World: the spatial objects and scenes of physical reality.
Optical World: the changing appearance conveyed towards an eye, camera or detector through emitted or reflected light.
Visual World: the transformation and interpretation of this optical information through human physiological and psychological vision.
Represented World: corresponding spatial information embodied in drawings, photographs, films, computer images and other representations.

Visual perception therefore occupies a crucial position between physical reality and the experienced visual world.

The Main Components of Visual Perception

The Dictionary identifies eight important functional components of visual perception:

Visual Discrimination;
Visual Memory;
Visual Spatial Relations;
Figure–Ground;
Visual Closure;
Form Constancy;
Visual Tracking; and
Visual Motor Integration.

These functions demonstrate that seeing is not simply passive image reception. Perception requires identification, comparison, organisation, recognition, memory, spatial judgement and interaction.

Visual Discrimination

Visual Discrimination is the ability to identify similarities and differences between visible features.

It allows the observer to distinguish differences of size, shape, orientation, colour, texture and other visual characteristics.

This is fundamental to perspective because interpreting a spatial scene requires comparison between Forms and between the transformed appearances of similar Forms at different positions or distances.

Visual Memory

Visual Memory enables previously observed visual information to contribute to current perception.

Objects do not need to be interpreted as entirely unfamiliar Forms on every encounter. Earlier experience helps the observer recognise expected sizes, shapes, orientations and relationships.

Knowledge that a door, person, vehicle or building possesses a familiar physical Form can therefore help resolve otherwise ambiguous projected appearances.

Visual Spatial Relations

Visual Spatial Relations concern the perceived positions and relationships of objects and Forms within space.

These include judgements such as above and below, left and right, near and far, inside and outside, in front and behind, parallel and angled, attached and separated.

Perspective phenomena provide much of the changing geometrical information from which such spatial relationships can be interpreted.

Figure–Ground Perception

Figure–Ground perception allows the visual system to distinguish an object or Form from its surrounding visual field.

A complex retinal image contains numerous edges, colours, textures, shadows and overlapping regions. The observer must organise these into meaningful objects and background regions rather than experiencing the image as an undifferentiated pattern.

Edges, contrast, occlusion and contextual organisation all contribute to this segregation of visual Forms.

Visual Closure

Visual Closure enables an object to be recognised even when part of its visible Form is missing, hidden or interrupted.

An object partly concealed behind another object is not normally perceived as physically truncated at the occluding edge. The visual system can infer or complete the likely continuation of its Form.

This capacity is essential in normal spatial vision because occlusion is ubiquitous in complex three-dimensional scenes.

Perceptual Constancy

Perceptual constancy describes the tendency to perceive an object as remaining comparatively stable even though its immediate sensory appearance changes.

The Dictionary identifies several related constancies, including:

Size Constancy: an object tends to be perceived as maintaining its physical size despite changes in viewing distance and retinal image size.
Shape Constancy: an object can be recognised as maintaining its Form despite changes in viewing angle and projected shape.
Brightness Constancy: apparent object brightness can remain comparatively stable under different illumination.
Colour Constancy: perceived object colour can remain comparatively stable under changing lighting.
Location Constancy: an object can be perceived as occupying a stable environmental position despite changing retinal relationships.

Perceptual constancy is particularly important to perspective because perspective continuously transforms retinal size, projected shape and apparent position as viewpoint and distance change.

Size Constancy and Perspective

The retinal image of an object becomes smaller as its viewing distance increases, yet the observer does not normally experience a familiar person as physically shrinking while walking away.

Visual Size Constancy helps maintain an interpretation of stable physical size by combining retinal size with information concerning distance and surrounding context.

This distinction between projected or retinal size and perceived physical size is fundamental to understanding the relationship between visual perspective and spatial perception.

Shape Constancy and Aspect

A similar principle applies to shape.

A rectangular surface can produce many different projected Forms as it rotates relative to the observer. Nevertheless, the visual system may continue to identify the underlying object as the same rectangle.

Shape or Form Constancy therefore operates alongside Aspect Perspective. Perspective explains why the image changes; perceptual constancy helps explain why the object itself can nevertheless be experienced as stable.

Visual Tracking

Visual Tracking involves moving the eyes in order to follow a path, object or changing visual target.

Tracking allows the point of fixation to remain related to a moving Form and consequently alters the relationship between the fixed retinal field, eye orientation and surrounding spatial environment.

Natural visual perception is consequently dynamic and active rather than equivalent to the permanently fixed viewpoint assumed by a single static perspective drawing.

Visual Motor Integration

Visual Motor Integration concerns the coordination of visual information with bodily action.

Seeing is used to guide reaching, grasping, walking, turning, navigating and countless other interactions with spatial reality.

Accurate interpretation of distance, direction, size, shape and position therefore has practical consequences. Visual perspective assists not merely in recognising the world but also in acting within it.

Perceiving Perspective Images

Visual perception is required not only for viewing physical spatial reality but also for interpreting representations of spatial reality.

A perspective drawing, photograph, film or computer image is physically a two-dimensional or otherwise limited representation, yet the visual system can interpret its internal geometrical and optical cues as indicating an extended three-dimensional scene.

The observer therefore simultaneously encounters:

the physical image surface; and
the represented spatial world perceived within or beyond that surface.

Understanding picture perception is consequently central to the relationship between Visual Perspective Type 2 and graphical, photographic, cinematic and digital perspective.

Perspective Images and the Illusion of 3-D

A flat perspective image can evoke powerful depth because it preserves or reproduces selected monocular cues found in natural vision.

Converging parallel directions, diminution of size, foreshortening, occlusion, texture gradients, atmospheric effects, shadows and familiar Forms can all contribute to the interpretation of a three-dimensional represented space.

Stereoscopic information is therefore not necessary for every effective representation of depth.

The perceptual system can construct a strong impression of spatial extension from a single fixed perspective image.

Perspective Illusions

The same perceptual mechanisms that normally help us interpret spatial reality can sometimes produce visual or perspective illusions.

The Dictionary distinguishes broad illusion sources including:

Physical Illusions: arising from physical or environmental conditions;
Physiological Illusions: arising from characteristics of the eye and visual system; and
Cognitive Visual Illusions: arising from perceptual interpretation and inference.

Perspective illusions demonstrate that perception does not always correspond perfectly with physical geometry.

Context, apparent depth, relative size, surrounding Forms and other visual cues can alter the perceived size, shape, position or orientation of otherwise unchanged image elements.

Perceived Visual Angle

The physical visual angle subtended by an object is a geometrical relationship between object size and viewing distance.

The perceived visual angle, however, may not always correspond exactly to this physical angle. Apparent distance, context, surrounding objects and other depth cues can influence the subjective appearance of size.

This distinction again illustrates why retinal or projection geometry alone does not provide a complete theory of visual perception.

Perspective Phenomena and Human Perception

The PRC uses the term Perspective Phenomena for the apparent and generalised changes to visual features associated with particular perspective processes, views and representations.

For human perception, important examples include:

Aspect of View;
Diminution of Size;
Diminution of Form;
Degradation of Form;
Aspect Foreshortening;
Perspectival Foreshortening;
converging parallel directions and vanishing points;
horizon relationships;
diminution of colour and contrast;
gradients of visual acuity;
texture;
motion parallax;
occlusion;
light and shade; and
binocular disparity and vergence.

These phenomena provide much of the information through which visual perception identifies the apparent Form and spatial arrangement of objects and scenes.

Environmental, Extrinsic and Intrinsic Factors

The Dictionary also provides a broader classification of Perspective Phenomena that is useful for visual perception.

Environmental factors: contextual and ecological relationships associated with the observer and surrounding environment.
Extrinsic factors: overt transformations originating in spatial or optical reality, including changes associated with light-ray direction, intensity, colour, Form and distance.
Intrinsic factors: effects associated with the observation or imaging system itself, including monocular and binocular optical processes.

Human perception may therefore result from several different sources operating together rather than from one isolated visual mechanism.

Physiological and Psychological Optics

Visual perception lies at the intersection of physiological optics and psychological optics.

Physiological processes include the functioning of the eyes, retina, photoreceptors, focusing mechanisms, eye movements and associated nervous pathways.

Psychological processes concern the interpretation, organisation and experienced meaning of the resulting visual information.

The distinction is analytical rather than absolute. Normal perception arises from their interaction.

Visual Perception Is Contextual

The same retinal or image feature can be interpreted differently according to its surrounding visual context.

Apparent size depends partly upon inferred distance. Apparent orientation can be influenced by surrounding lines. Apparent shape can depend upon the interpretation of depth. A partially visible edge may be understood as belonging to a complete hidden object because of surrounding structural information.

Visual interpretation consequently involves relationships among image elements rather than the isolated reading of individual points or pixels.

Visual Perception Is Active

Natural seeing is an active process.

The observer can change fixation, move the eyes, turn the head, alter viewing distance, walk around an object and approach or retreat from a spatial scene.

Every movement changes aspects of the available perspective information. Ambiguous Forms can become clearer, hidden surfaces may become visible and motion parallax can reveal depth relationships that were uncertain from the previous viewpoint.

Visual perception therefore operates across changing viewpoints and through time rather than being permanently restricted to the geometry of one instantaneous projection.

Visual Perception and Orientation

Vision also provides essential information for spatial orientation.

Perspective relationships help establish the apparent directions of lines and planes, the orientation of surfaces, the position of the ground plane, the distribution of objects and the observer’s changing relationship to surrounding space.

Movement, horizon relationships, optic flow, vanishing structures, occlusion and environmental invariants can all contribute to determining where the observer is and how the surrounding environment is organised.

Visual Perception and Navigation

The practical function of visual perception extends directly to navigation through spatial reality.

Walking through a room, crossing a road, reaching for an object or moving around an obstacle all require continuous interpretation of relative direction, distance, scale, position and movement.

The many phenomena grouped under perspective therefore have practical behavioural functions. They allow spatial structures to be visually modelled and acted upon rather than merely admired as appearances.

Visual Perception and Graphical Perspective

Graphical perspective is effective partly because it reproduces selected visual relationships that the human perceptual system already uses when interpreting physical space.

A linear-perspective drawing may reproduce diminution of size, convergence, vanishing points, horizon relationships, foreshortening, occlusion and other monocular cues.

The observer interprets these marks on a flat surface as corresponding to an extended represented spatial scene.

Graphical perspective therefore depends upon both image construction and human visual interpretation.

Visual Perception and Photography

A photograph provides an optical perspective image captured from a particular camera position, direction and field of view.

When the photograph is subsequently viewed, Visual Perspective Type 2 operates again as the observer interprets the photographic image.

There are consequently at least two relevant stages: formation of the camera image and perception of that image by a human viewer.

The camera and eye are related optical systems, but the camera image is not a complete physical duplicate of normal human visual experience. Viewing conditions, field of view, image size, projection geometry and perceptual processing all contribute to the final appearance.

Visual Perception and Moving Images

Film, television and other moving-image systems add changes through time to the monocular cues already present in still perspective images.

Camera movement, changing viewpoint, occlusion, motion parallax, changing scale and moving reflections or shadows can provide additional information about spatial structure.

Moving-image perspective can therefore produce particularly strong impressions of spatial reality even when the display itself remains physically flat.

Visual Perception in Virtual Reality

Virtual Reality can combine many of the visual cues used in natural spatial perception.

Depending upon the system, these may include monocular perspective, binocular disparity, vergence relationships, changing viewpoint, motion parallax, wide field of view, occlusion and interactive movement through a modelled environment.

Such systems demonstrate that the experienced perception of three-dimensional space depends upon the complete relationship between image generation, display, observation mode and the human visual system.

Visual Perception and Spatial Realism

An image does not need to reproduce every characteristic of natural vision in order to produce a compelling sense of spatial reality.

A relatively small subset of consistent monocular cues can be sufficient for the visual system to interpret a represented scene as three-dimensional.

This helps explain the success of drawings, paintings, photographs and conventional cinema, all of which omit many aspects of direct natural viewing but can nevertheless provide convincing spatial impressions.

Visual Perception and Information Loss

Every visual view is limited.

A single viewpoint conceals surfaces behind other surfaces, reduces distant details, transforms apparent shapes and provides only a selected directional sample of the spatial scene.

Additional information can also be lost through limited visual acuity, low contrast, atmospheric effects, restricted field of view and the finite resolution of optical or represented images.

Visual perception must therefore operate upon incomplete information and infer aspects of spatial reality that are not directly contained within one retinal or represented image.

Visual Perception Is Not Fully Understood

The Dictionary and Volume 1 both emphasise that visual perception remains an exceptionally complex field.

Much is known about retinal imaging, visual acuity, depth cues, eye movement, binocular disparity and other mechanisms, but the complete operation through which physiological and psychological processes generate the stable and detailed visual world remains incompletely understood.

Perspective theory should therefore distinguish well-established geometrical and optical relationships from the more complex processes through which those relationships are ultimately perceived and interpreted.

Common Misconceptions about Visual Perception

Several common assumptions should be avoided:

Visual perception is not simply the retinal image. The retinal image is an optical input to a much more complex perceptual process.
The eye does not directly measure absolute depth from a single retinal direction. Depth must be inferred from additional information.
Binocular stereopsis is not the only source of three-dimensional perception. Monocular depth cues provide powerful spatial information.
A flat image does not need to be stereoscopic to communicate 3-D space. Perspective drawings, photographs and ordinary films demonstrate this clearly.
The visual field is not uniformly sharp. Central acuity is much greater than peripheral acuity.
The visual field and visual world are not the same concept. One describes the current bounded field; the other the stable extended environment perceived across changing views.
Perceived size is not identical to retinal image size. Size constancy and depth interpretation can stabilise apparent physical size.
Perceived shape is not identical to instantaneous projected shape. Form constancy helps preserve object identity across changes of aspect.
Natural vision is not normally a fixed-viewpoint process. The eyes, head and observer continually move.
Perspective illusions do not prove that perception is generally unreliable. They reveal circumstances in which ordinarily useful perceptual processes produce interpretations that differ from physical geometry.

Why Visual Perception Matters

Visual perception is fundamental to perspective because perspective images and views have no practical visual meaning until their information is received and interpreted by an observer.

The human visual system transforms optical information into an organised experience of Form, size, shape, position, orientation, movement, distance and depth. It does this by combining retinal image structure with monocular, binocular, oculomotor, environmental, contextual and perceptual information.

This explains why the same underlying principles link the direct perception of physical reality with the interpretation of drawings, paintings, photographs, cinema, computer graphics and Virtual Reality. All ultimately depend upon the visual system recognising and interpreting perspective phenomena.

Understanding Visual Perception therefore provides a foundation for understanding Visual Perspective Type 2, Optics of the Eye, physiological optics, psychological optics, Spatial Perception, depth cues, visual field, Field of View, Visual Space, Visual World, visual acuity, Perspective of Form, diminution of size, foreshortening, occlusion, motion parallax, optic flow, stereopsis, binocular vision, perceptual constancy, perspective illusions and the correspondence or equivalence problem.