Computer Perspective

Computer Perspective refers to perspective views, images, models and spatial representations generated, processed, analysed, manipulated, displayed or explored through computer systems. It extends perspective from traditional graphical and optical methods into computational forms of spatial modelling, visualisation and interaction.

The Dictionary groups Computer / Digital Perspective within the wider field of New Media Perspective. Important applications include computer graphics, Computer-Aided Design (CAD), Computer-Generated Imagery (CGI), computer and robotic vision, ray tracing, digital imaging, computer games, Virtual Reality, Augmented Reality, artificial intelligence and interactive three-dimensional environments.

Computer Perspective is therefore much broader than simply drawing a conventional perspective image with a computer. A computer can construct a three-dimensional spatial model, calculate views from changing viewpoints, apply different projection systems, analyse existing images, simulate light and optical effects, reconstruct spatial reality and allow users to move interactively through virtual space.

Computer Perspective and Digital Perspective

The Dictionary frequently combines the terms Computer Perspective and Digital Perspective, reflecting their extensive overlap.

Digital Perspective is defined more generally as a digital perspective view, image or model of spatial reality created from, or projected from, physical, instrument, imaginary, mathematical, graphical or New Media sources.

Computer Perspective refers particularly to the computational methods, systems and models through which such digital spatial views can be generated, transformed, analysed or explored.

The two terms therefore belong to the same closely connected field and may frequently describe different aspects of the same perspective process.

Computer Perspective and New Media Perspective

New Media Perspective is the broader Perspective Category Theory category for perspective generated, processed, displayed or explored through computational and electronic media.

It includes:

computer graphics;
digital imaging;
computer vision;
artificial intelligence;
Virtual Reality;
Augmented Reality;
Mixed Reality;
Extended Reality;
interactive 3-D environments;
networked visual systems; and
other computational forms of spatial representation.

Computer Perspective is therefore one important area within the much wider development of New Media Perspective.

From Manual Perspective to Computational Perspective

Traditional graphical perspective requires the artist or draughtsperson to construct the projected relationships of a spatial scene manually.

Computer systems can instead store the geometry of an object or environment mathematically and automatically calculate corresponding perspective views.

This development greatly expands the number, complexity and variety of possible views. A three-dimensional model can be examined from changing positions, orientations, scales and Fields of View without reconstructing the underlying object for each new image.

Perspective consequently becomes not merely a fixed picture but a dynamic system for modelling and exploring spatial reality.

The 3-D Computer Model

A fundamental component of Computer Perspective is the three-dimensional computer model.

The Dictionary describes 3-D modelling as the computer-graphics process of developing a mathematical coordinate-based representation of the visible surfaces of a spatial object in three dimensions.

Edges, vertices, polygons and related geometrical structures can be manipulated within a digital 3-D image or model space.

Once the spatial structure exists as a computer model, many different perspective images can be generated from it.

Computer Reality

The Dictionary uses Computer Reality for a spatial reality generated within a digital model.

Such a model may represent an existing physical scene, a proposed object or building, a reconstructed environment, an entirely imaginary world or a combination of several sources.

This provides an important distinction between physical object space and the computational or digital space within which Computer Perspective operates.

Virtual Perspective Models

Volume 1 defines a Perspective Model as a system for constructing a comprehensive three-dimensional visual representation of a spatial reality.

A Virtual Perspective Model is a computer-graphics model representing the overall Form and structure of a spatial reality together with relevant mathematical, material or optical processes.

Examples include:

CAD models;
CGI models;
GIS environments;
computer-vision reconstructions;
Virtual Reality environments;
Augmented Reality systems;
Mixed Reality systems; and
Extended Reality systems.

Such models can contain linked views at different positions, scales and times and can often be explored interactively.

Computer-Aided Design — CAD Perspective

Computer-Aided Design (CAD) is one of the major applications of Computer Perspective.

The Dictionary describes CAD drawing as a wide-ranging set of New Media methods for producing perspective images of real or imaginary spatial objects during the design process.

CAD can employ traditional technical views such as:

plans;
elevations;
orthographic projections;
parallel perspective;
axonometric views; and
linear-perspective views.

Unlike a static hand drawing, a CAD model can automatically generate new views as the object, viewpoint or projection conditions are changed.

Computer-Generated Imagery — CGI

Computer-Generated Imagery (CGI) is another major application of Computer Perspective.

CGI enables spatial objects, scenes, characters and environments to be created within a computer model and subsequently represented as still or moving perspective images.

The underlying computer model can contain geometry, viewpoint information, projection relationships, lighting, surface appearance and other visual features needed to construct a represented spatial world.

CGI may reproduce familiar perspective forms or create viewpoints and environments that would be difficult or impossible to obtain through ordinary photography.

Wire-Frame Perspective

Wire-Frame Perspective is a common Computer Perspective representation associated with CGI and CAD.

A digital 3-D model is displayed primarily through its structural edges or lines, giving the appearance of a spatial object constructed from a wire framework.

This representation allows the underlying geometrical structure to be built, inspected and changed before more detailed surfaces and visual effects are added.

Wire-frame representation demonstrates particularly clearly the relationship between traditional perspective geometry and modern computational modelling.

Computer Perspective Projection

A computer model does not possess only one possible perspective projection.

Different projection rules can be applied according to the intended purpose of the image or model. These can include familiar perspective forms such as:

central projection;
linear or rectilinear perspective;
parallel projection;
orthographic perspective;
axonometric perspective;
oblique perspective;
curvilinear perspective; and
spherical or panoramic forms.

Computer Perspective therefore does not replace older perspective systems. It provides computational means for generating, combining, transforming and exploring them.

The Virtual Camera

Many computer perspective systems employ a virtual camera to establish a view within the computer model.

As with a physical eye or camera, important variables include:

viewpoint or camera position;
Direction of Vision or Line of Sight;
Field of View;
orientation;
projection geometry; and
the spatial relationship between camera and model.

Changing these parameters generates a different perspective view of the same digital spatial reality.

Changing Viewpoint

One of the most important advantages of a 3-D computer model is the ability to generate many different views of the same Form.

The virtual viewpoint can move around, towards, away from, above, below or within the model. Each change of position or Direction of Vision produces corresponding changes of apparent size, shape, orientation, overlap and spatial relationship.

The computer can therefore demonstrate perspective as a continuously changing relationship between observer and spatial reality rather than merely as one completed image.

Unlimited Viewing Angles

The Dictionary associates digital computer modelling with unlimited viewing-angle representation.

A spatial model can potentially be explored from an extremely large number of positions and directions rather than being restricted to the single fixed view contained in an ordinary drawing or photograph.

This provides the basis for interactive CAD models, virtual worlds and other systems in which the observer can move freely through or around a represented spatial environment.

Computer Perspective and Vanishing Points

Computer systems can calculate the directional geometry of complex scenes automatically.

A digital linear-perspective image is therefore not restricted to the simplified one-, two- or three-point constructions commonly taught in elementary drawing.

Complex three-dimensional scenes can contain numerous families of spatial parallels and therefore numerous corresponding vanishing points. Computer-generated perspective can calculate these relationships directly from the stored spatial geometry and virtual viewpoint.

Volume 1 consequently recognises multi-point and unlimited-point linear perspective as normal possibilities within complex computer-generated imagery.

Computer Ray Tracing

Computer Ray Tracing is a major Computer Perspective method for modelling the interaction of light with a three-dimensional scene.

The Dictionary distinguishes two broad uses:

Computer Rendering: mathematical simulation of light within a computer model in order to generate an image.
Optical Design: computer modelling and characterisation of physical optical systems.

In rendered spatial scenes, ray tracing can simulate effects such as reflection, refraction, shadows and indirect lighting.

It therefore extends Computer Perspective beyond the geometry of lines and surfaces into the optical appearance of represented spatial reality.

Rendering Computer Perspective

A computer model contains geometrical and other spatial information, but this information must be transformed into a visible image before it can be viewed on a conventional display.

Rendering generates the represented view of the digital spatial model according to the selected viewpoint, projection and visual conditions.

The resulting image can range from a simple structural or wire-frame representation to a complex visual image incorporating surface appearance, lighting, shadows, reflections and other optical effects.

Computer Animation

Computer Perspective can also change through time.

Computer-generated animation creates sequences of images in which spatial objects, viewpoints or both can change between frames.

This allows perspective phenomena such as changing apparent size, shape, position, orientation, occlusion and viewpoint to be represented continuously.

Computer animation therefore combines spatial perspective with Motion Perspective.

Computer Games

Computer games provide one of the most familiar interactive applications of Computer Perspective.

Volume 1 describes how the development of 3-D accelerated graphics enabled games to move beyond simple two-dimensional sprites towards rendered spatial worlds employing 3-D geometry, perspective projection and ray tracing.

Modern game environments can allow the viewpoint to move continuously through a digital world, producing a new perspective image in real time as the player’s position and direction change.

First-Person Perspective

In First-Person Perspective, the game view is presented approximately from the viewpoint of the player-controlled character.

The visible world therefore appears directly ahead from that character’s position, producing a strong correspondence between the player’s view and the virtual viewpoint within the digital environment.

Second-Person Perspective

Second-Person Perspective is a rarer and more disputed game-view arrangement in which the player-controlled character can be seen through the viewpoint of another character, object or external agent within the game world.

The term concerns the visual viewpoint and should not be defined merely by second-person narration.

Third-Person Perspective

In Third-Person Perspective, the player’s avatar is visible within the represented image and the virtual camera is normally positioned outside or behind the character.

The player therefore observes both the represented environment and the character occupying it from an external virtual viewpoint.

Computer Vision

Computer Vision represents another major branch of Computer Perspective, but its direction is different from ordinary CGI.

Rather than primarily generating a new image of a computer model, computer vision analyses visual data in order to identify and interpret objects, scenes and processes contained within images or video.

Volume 1 distinguishes broad computer-vision functions including:

object and process detection;
image classification;
facial recognition;
spatial awareness;
navigation;
robotic vision;
database-image analysis; and
interpretation of visual information for decision-making.

Computer Vision therefore demonstrates that Computer Perspective concerns not only the production of perspective images but also their analysis and interpretation.

Robotic Vision

Robotic Vision is a specialised application of computer vision involving live visual sensing, spatial interpretation and associated real-time action.

A robot or autonomous system can use camera or sensor information to identify objects, establish spatial relationships, navigate an environment and perform tasks.

This extends perspective from human observation and image representation into automated machine perception and interaction with spatial reality.

Artificial Intelligence and Computer Perspective

Volume 1 identifies Artificial Intelligence as an increasingly important part of contemporary perspective technology.

Possible and existing functions include:

generation of perspective images;
generation of novel viewpoints;
image identification and classification;
analysis of perspective views;
reconstruction or reverse-engineering of scene geometry;
navigation through linked images;
integration of multiple images into larger image spaces; and
robotic interpretation of visual scenes.

Artificial Intelligence therefore extends both the image-generating and image-interpreting functions of Computer Perspective.

Novel View Generation

A particularly important computer function is the generation of a new perspective viewpoint from existing spatial information.

If sufficient information about the underlying scene or model is available, a computer system can generate views from positions different from those originally captured or displayed.

This ability connects computer graphics, computer vision, 3-D reconstruction and artificial intelligence and represents a major extension beyond the fixed viewpoint of an ordinary photograph.

Computer Reconstruction of Spatial Reality

Computer systems can also construct digital spatial models from information captured from physical reality.

Images from different viewpoints can be compared and related to infer the position and Form of spatial structures. The resulting information may be represented as geometrical models, point clouds or other forms of digital spatial representation.

The process reverses the ordinary direction of image formation: perspective images become evidence from which aspects of the underlying spatial reality can be reconstructed.

Virtual Reality

Virtual Reality is an interactive computer-generated, reconstructed or captured environment that can be explored from changing viewpoints.

Contemporary systems may combine computer-generated perspective with stereoscopic imagery, head or body tracking and other forms of interaction.

As the user turns or moves, the represented viewpoint changes correspondingly. The perspective image is therefore generated dynamically rather than remaining fixed.

Virtual Reality is an important example of Computer Perspective developing from image representation towards immersion within a represented spatial environment.

Augmented Reality

Augmented Reality combines digitally generated visual information with a view of physical spatial reality.

An AR system can capture the physical environment, analyse its geometry or position and overlay virtual information or Forms upon the resulting view.

Computer Perspective is required to coordinate the viewpoint, scale, position and orientation of the digital elements with the physical environment in which they appear.

Mixed and Extended Reality

Mixed Reality and Extended Reality further combine physical, digital, simulated and interactive spatial environments.

Extended Reality is used as an umbrella term for Virtual, Augmented and Mixed Reality systems.

These systems illustrate how Computer Perspective can participate in a larger perspective process containing several categories simultaneously.

Digital Twins and Virtual Worlds

Computer models can be used to create digital representations of physical or imagined environments at many scales.

Such environments may form digital twins, virtual worlds, networked 3-D spaces or other forms of computer reality.

The same underlying model can potentially be inspected from multiple viewpoints, displayed at different scales and connected to changing information about the represented spatial reality.

Digital Metaverse and Multiverse

The Dictionary also identifies the concepts of a Digital Metaverse and Digital Multiverse.

A Digital Metaverse concerns persistent or interconnected virtual environments in which users and digital objects may interact.

A Digital Multiverse concerns multiple interconnected or coexisting virtual worlds, environments or simulations, potentially combined with physical reality through VR, AR, MR or XR systems.

These concepts extend Computer Perspective beyond the production of individual images towards navigable and interconnected spatial worlds.

Virtual Production

Virtual Production combines computer-generated spatial environments with physical filmmaking.

Digital scenes can be displayed on large LED surfaces while camera tracking updates the represented perspective and lighting according to the physical camera’s movement.

The process can combine:

physical architecture or actors;
Camera Perspective;
computer-generated imagery;
real-time rendering;
camera tracking;
digital environments; and
New Media display systems.

Virtual Production is therefore a clear example of several perspective categories operating together.

Computer Perspective and Category Chaining

Perspective Category Theory is particularly useful for understanding computer-generated and digitally processed images because they rarely arise from one isolated process.

A practical image chain might involve:

Natural scene → Optical and Instrument Perspective → digital or New Media processing → computer-generated or processed image → New Media or Instrument display → Visual Perspective Type 2.

When perspective categories operate successively in this way, the Dictionary describes the process as Category Chaining.

A modern perspective image may consequently have physical, optical, instrument, mathematical, graphical, computational and perceptual stages within one complete system.

Category Overloading

A computer-generated perspective image can also belong legitimately to several perspective categories at the same time.

A digitally rendered linear-perspective image, for example, can be:

Mathematical Perspective because geometrical or algorithmic rules govern the projection;
Graphical Perspective because a represented spatial image is produced; and
New Media Perspective because the image is computationally generated and displayed.

Perspective Category Theory describes such legitimate overlap as Category Overloading.

Computer Perspective and Composite Perspective

Computer methods also make it easy to combine several kinds of perspective within one resulting image or environment.

A film scene might combine real architecture, forced perspective, optical camera imaging, digital compositing and CGI. A virtual or augmented environment may combine captured physical imagery with reconstructed or entirely generated Forms.

The resulting perspective can therefore be a Composite Perspective produced through several interacting perspective methods or categories.

Computer Perspective and Mathematical Perspective

Computer Perspective depends extensively upon Mathematical Perspective.

Volume 1 describes Mathematical Perspective as the application of algorithmic rules to transform or model the appearance of scene and object geometry.

Important variables can include:

scale;
depth and distance;
Field of View;
viewpoint;
shape and aspect;
colour;
spatial resolution; and
projection geometry.

Computer systems provide powerful means of applying such mathematical relationships to complex spatial models.

Computer Perspective and Graphical Perspective

A computer-generated perspective image may reproduce many of the geometrical principles traditionally associated with Graphical Perspective.

These can include apparent diminution of size, aspect changes, foreshortening, convergence of spatial parallels, vanishing points, horizon relationships and projected shape.

The difference is that the geometrical relationships can be calculated automatically from a digital spatial model rather than constructed individually by hand.

Computer Perspective Is Not Necessarily Linear Perspective

Computer Perspective should not be equated with Linear Perspective.

A computer can generate a rectilinear central-perspective image, but it can also employ parallel, curvilinear, cylindrical, spherical, panoramic, multi-view or other projection systems.

Linear Perspective is therefore one possible geometrical form that can operate within Computer Perspective, not the definition of Computer Perspective itself.

Computer Perspective Is Not Necessarily Photorealistic

Computer Perspective can be used for many purposes other than producing a photograph-like image.

A computer model may be represented as:

a wire-frame model;
a technical drawing;
a plan or elevation;
a parallel projection;
a rendered perspective image;
a scientific visualisation;
a navigable virtual environment; or
another mathematical or graphical representation.

The required perspective form depends upon the purpose of the model rather than upon a universal requirement for visual realism.

Perspective as Method and Outcome

Computer Perspective illustrates the Dictionary’s broader principle that:

Perspective = Method or Process + Resulting Image, View or Spatial Appearance.

The term Computer Perspective can therefore refer both to computational methods for generating or analysing spatial views and to the perspective images, models or environments produced by those methods.

Functions of Computer Perspective

Volume 1 identifies several general functions or goals of perspective that are particularly relevant to computer systems:

View — observe or capture spatial reality;
Match — measure, compare, classify or cross-match spatial information;
Represent — model, copy, index, link or explore spatial reality;
Illusion — generate a convincing alternative spatial appearance; and
Immersion — create the impression of being present within a represented spatial environment.

Computer Perspective can contribute to every one of these functions.

Computer Perspective in Science and Engineering

Computer Perspective is also important for spatial problems in science, engineering and technical practice.

Volume 1 associates computational perspective methods with fields including:

Computer-Aided Design;
engineering;
architecture;
Geographical Information Systems;
cartography;
photogrammetry;
medical imaging;
computer vision;
robotics;
remote sensing;
navigation;
astronomy; and
scientific modelling.

In these fields, perspective can function not simply as a means of making pictures but as a method of modelling, measuring, analysing and manipulating spatial information.

Computer Perspective and the Optical Image Chain

A digital perspective image may form only one stage within a much longer Optical Image Chain.

A physical scene may first be captured by a camera, converted into digital information, computationally processed, combined with CGI, displayed on a screen and finally viewed by the human visual system.

Perspective can therefore change category and form as an image passes through successive physical, optical, instrument, digital and perceptual stages.

Computer Perspective and the Future of Perspective

Volume 1 treats computational media as one of the principal areas in which perspective continues to expand.

Computer graphics, computer vision, CGI, CAD, artificial intelligence, Virtual Reality, Augmented Reality, robotics, digital filmmaking, virtual production and related technologies are extending the ways spatial reality can be captured, modelled, analysed, represented and explored.

The significance of Computer Perspective is therefore not simply that computers make traditional perspective construction faster. Computational systems allow entirely new relationships between spatial models, images, viewpoints, measurements, simulations, users and represented realities.

Common Misconceptions about Computer Perspective

Computer Perspective is not simply perspective drawing performed on a computer. It includes modelling, rendering, analysis, reconstruction, vision, interaction and immersive environments.
Computer Perspective is not synonymous with CGI. CGI is one important application within a much larger computational field.
Computer Perspective is not synonymous with Linear Perspective. Computers can generate linear, parallel, curvilinear, spherical and many other projection forms.
A computer model is not restricted to one viewpoint. It can generate many or potentially unlimited viewing positions and directions.
Computer Perspective does not require photorealism. Technical, wire-frame, mathematical and abstract representations are equally valid computational perspective products.
Computer Perspective is not restricted to image generation. Computer Vision and Robotic Vision analyse and interpret perspective images.
Digital Perspective and Computer Perspective overlap strongly but are not necessarily the only terms required. The Dictionary places both within the broader context of New Media Perspective.
Perspective principles have not been made obsolete by computers. Computers calculate and apply geometrical, mathematical, optical and visual relationships rather than eliminating them.
One-, two- and three-point perspective do not exhaust computer-generated linear perspective. Complex scenes may contain many or unlimited directional vanishing points.
A digital perspective image may belong to several categories simultaneously. Mathematical, Graphical and New Media Perspective can operate together.
A computer-generated image may form only one stage within a larger perspective chain. Physical, optical, instrument, computational, display and perceptual processes can all contribute to the final viewed image.

Why Computer Perspective Matters

Computer Perspective matters because computers have transformed perspective from a predominantly fixed system of drawing or imaging into a dynamic means of modelling, generating, analysing, reconstructing, navigating and interacting with spatial reality.

A single digital model can generate many views, Fields of View, scales, projection types and visual conditions. It can represent an existing physical world, assist in the design of a future object, reconstruct a scene from images, create an imaginary environment or respond continuously to a moving observer.

Computer Perspective therefore connects traditional perspective geometry with some of the most important contemporary visual technologies, including CAD, CGI, Computer Graphics, Computer Vision, Robotic Vision, ray tracing, artificial intelligence, computer games, Virtual Reality, Augmented Reality, Extended Reality, GIS, digital filmmaking and virtual production.

Understanding Computer Perspective provides a foundation for understanding Digital Perspective, New Media Perspective, Mathematical Perspective, Graphical Perspective, Computer-Aided Design, Computer-Generated Imagery, 3-D Modelling, Wire-Frame Perspective, Ray-Tracing Perspective, virtual cameras, Computer Vision, Robotic Vision, Artificial Intelligence, novel viewpoint generation, Virtual Reality, Augmented Reality, Mixed Reality, Extended Reality, Digital Reality, Computer Reality, Composite Perspective, Category Chaining and the modern computational representation of spatial reality.

Explore Theory

Explore the principal theories, classifications, types, forms, geometries, spatial concepts and visual phenomena of perspective.