Camera View / Image

A Camera View / Image is a perspective view formed by a camera, or the corresponding image captured, recorded or transmitted by that camera. The distinction between view and image is important: a Camera View may exist live or in real time before recording, whereas a Camera Image is the fixed or recorded visual product produced by the imaging process.

A camera transforms information from three-dimensional object space into an image in camera image space. The resulting view or image depends upon camera position, viewing direction, projection geometry, field of view, image format, focus, resolution, exposure, motion and optical characteristics.

Camera View / Image therefore concerns much more than photography. It includes still photography, cinematography, video, digital imaging, scientific cameras, robotic and computer vision, satellite and remote-sensing imagery, stereoscopic cameras, panoramic imaging and other systems in which cameras form views of spatial reality.

Camera View and Camera Image

The terms Camera View and Camera Image refer to closely related but distinguishable stages of camera imaging.

Camera View — the live or real-time view formed, displayed or available through the camera system.
Camera Image — the captured, recorded or stored image produced from that view.

A digital camera, for example, can continuously form a live electronic view before the shutter is released. Once a particular frame is captured and stored, that frame becomes a Camera Image.

Similarly, a cinema or video camera continually forms changing camera views while recording a succession of individual frames. The resulting moving image is therefore constructed from a temporal sequence of captured Camera Images.

A Captured Perspective Image

Within Perspective Category Theory, a Camera Image is a principal example of a Captured Perspective Image.

Captured perspective operates from a spatial reality towards an imaging system and image space. Light or other physical information originating from, reflected by or transmitted through the scene is gathered by the camera and converted into an image.

This belongs principally to the Viewing or Imaging Class of perspective. It should be distinguished from the Projecting Class, in which an already formed image or light pattern is projected forwards onto a screen, surface, object or environment.

A cinema system can employ both directions sequentially:

scene → camera capture → recorded image → projector/display → spectator

The camera captures the image; the cinema projector or display subsequently presents it.

Explore Instrument Perspective →

The Camera Image Chain

A modern camera image is normally produced through an image chain containing several successive stages:

spatial object or scene → light → lens or aperture → image plane → film or sensor → image processing → storage/transmission → display → human or machine viewing

Each stage can affect the final appearance or information contained in the image.

The scene provides the physical object-space information.
The optical system selects and redirects light.
The image plane receives the optical projection.
The film or sensor records or converts the image information.
The processor may alter colour, contrast, sharpness, geometry or other image properties.
The storage or transmission system preserves or communicates the image data.
The display produces another physical image for viewing.
The eye and visual system finally interpret the displayed image.

The picture ultimately seen by a spectator may therefore be several transformations removed from the original physical scene.

The Camera Point or Viewpoint

The position from which a camera forms an image can be described as its Camera Point, viewpoint, station point or effective centre of projection.

This is one of the most important determinants of Camera Perspective. Moving the camera changes the geometrical relationship between the camera and every object in the scene.

Camera position determines such relationships as:

relative apparent size of near and far objects;
overlap and occlusion;
visible sides and surfaces;
aspect of Form;
foreshortening;
relative object position;
convergence and vanishing relationships;
foreground/background relationships.

Moving closer to a subject generally increases differences between near and far objects. Moving farther away reduces those differences.

Explore Camera Perspective →

Each Camera Frame Has a Viewpoint

A crucial property of conventional camera imaging is that each individual captured frame represents the scene from a particular camera position and viewing direction during its exposure or frame-integration interval.

A still photograph therefore records a spatial scene from one principal viewpoint at one interval of time.

A moving camera does not abolish this principle. Instead, successive movie or video frames can be captured from successively different camera positions or directions.

A moving-image sequence can therefore be understood as:

Viewpoint 1 → Frame 1
Viewpoint 2 → Frame 2
Viewpoint 3 → Frame 3
Viewpoint 4 → Frame 4 → …

When displayed rapidly in sequence, these separate frames produce the experience of continuous visual motion.

Camera Position versus Focal Length

One of the most common misunderstandings in photography concerns the relationship between camera position and lens focal length.

For an ordinary rectilinear camera system, changing focal length while leaving the camera in precisely the same position principally changes:

field of view;
framing;
image magnification;
the portion of the scene recorded.

It does not by itself change the fundamental projective relationships between objects contained within the common region of the scene.

A longer-focal-length image from the same camera position can therefore resemble a crop and enlargement of the central region of a shorter-focal-length image, although resolution, focus, depth of field and lens-specific behaviour may differ.

What often creates the familiar difference between a close wide-angle photograph and a distant telephoto photograph is that the photographer has changed camera position.

Changing viewpoint changes perspective geometry. Changing focal length at a fixed viewpoint principally changes field of view and image scale.

Field of View

The field of view is the angular region of object space captured or displayed by a camera system.

For a given sensor or film format:

a shorter focal length normally produces a wider field of view;
a longer focal length normally produces a narrower field of view.

Field of view determines how much of the surrounding spatial environment appears within the camera frame. It therefore plays an important role in composition, image scale and the amount of contextual information contained in the image.

Camera systems can range from narrow telephoto views to extremely wide-angle, fisheye, panoramic and omnidirectional views covering 180°, 360° or, through multiple lenses or computational stitching, almost an entire surrounding sphere.

Camera Orientation

Camera orientation is separate from camera position. A camera can remain at the same physical location while its viewing direction changes.

Three fundamental rotations are:

pan — rotation to the left or right;
tilt — rotation upwards or downwards;
roll — rotation around the forward viewing axis.

These operations alter the region and orientation of the scene contained within the image.

Tilt is particularly important in architectural photography. If a planar camera image surface is tilted relative to vertical parallel lines in the scene, those lines generally project towards a finite vanishing point and appear to converge.

The familiar appearance of converging building verticals is therefore primarily a consequence of camera orientation and projection geometry, not simply the use of a wide-angle lens.

The Image Plane

The camera forms an optical image upon an image plane, normally occupied by photographic film or an electronic sensor.

In the simplest central-projection model, rays from points in object space pass through an effective optical centre and intersect corresponding points on the camera image plane.

The relationship can be simplified as:

object-space point → centre of projection → image-space point

This is closely related geometrically to the classical perspective-window construction, although a physical camera normally forms its sensor or film image behind the effective optical centre and the optical system introduces additional physical characteristics absent from an ideal geometrical construction.

Camera Projection

Many conventional cameras approximate a form of central projection, but not every Camera Image uses the same geometrical mapping.

Important camera projection forms include:

rectilinear projection — straight scene lines are normally represented as straight image lines;
fisheye projection — very wide angular fields are compressed into a bounded image and many straight lines become curved;
cylindrical projection — commonly used for panoramic imaging;
spherical or equirectangular projection — represents an extended or complete surrounding angular field;
stereoscopic imaging — two or more separated camera views record binocular or multi-view information;
computational or composite imaging — several images may be combined into one larger or differently mapped image.

The geometry of a Camera Image therefore cannot be inferred from the word camera alone. The projection method must also be identified.

Still Camera Images

A still Camera Image records one principal spatial view during a limited period of exposure.

The resulting image freezes relationships that may have existed only temporarily:

object position;
viewpoint;
orientation;
lighting;
movement;
focus;
depth relationships;
occlusion;
visible Form.

The recorded photograph is therefore a fixed perspective product even though the physical scene from which it was formed may have been changing continuously.

Moving Camera Images and Cinematography

A motion-picture or video camera records a rapid sequence of Camera Images. Displaying these frames in temporal order produces a moving-image view.

Camera movement introduces changing perspective information through time. Important movements include:

pan;
tilt;
roll;
dolly forwards or backwards;
lateral tracking;
crane or jib movement;
pedestal movement;
handheld movement;
drone movement;
stabilised or motion-controlled movement.

Translation of the camera through object space changes the viewpoint and therefore changes perspective geometry. This creates dynamic phenomena including:

motion parallax;
changing overlap;
changing apparent size;
changing aspect and foreshortening;
revealed and concealed surfaces;
changing vanishing structures;
changing relationships between foreground and background.

Cinematography therefore adds Motion Perspective to the fixed spatial perspective contained within each individual frame.

Zoom is not Camera Movement

A lens zoom should be distinguished from physical camera movement.

When the camera remains stationary and focal length is changed, the image field and magnification change but the camera viewpoint remains fixed.

A dolly or tracking movement physically moves the camera and therefore changes perspective geometry.

A zoom changes focal length and field of view without necessarily changing viewpoint.

A dolly zoom combines both operations: the camera moves while focal length changes in the opposite direction. A chosen subject can remain approximately constant in image size while foreground/background perspective relationships visibly change.

Live View and Viewfinders

Before an image is recorded, many cameras provide a live Camera View through an optical or electronic viewfinder or display.

Different systems provide this view in different ways:

an optical viewfinder may provide a separate optical path;
a reflex camera can redirect light from the principal taking lens into the viewfinder;
a mirrorless digital camera reads the sensor continuously and displays an electronic live view;
a remote camera may transmit its live view to another screen or control centre;
a robotic camera may send the image directly to a computer rather than to a human observer.

The live view is therefore itself a perspective product, even before a permanent image has been recorded.

Film and Digital Camera Images

A photographic camera can record its image chemically or electronically.

Film photography forms an optical image upon a light-sensitive photographic emulsion. Chemical processing subsequently produces a persistent photographic image.

Digital photography forms the optical image upon an electronic sensor. The detected information is converted into digital data that can be processed, stored, transmitted and displayed.

Although the recording technologies differ, both systems can share essentially the same underlying viewpoint and projection geometry.

Image Format and Framing

The camera does not normally capture an unlimited optical field. A sensor, film frame or image-processing system selects a bounded portion of the available projection.

Important image-format properties include:

sensor or film dimensions;
aspect ratio;
image orientation;
pixel dimensions or film resolution;
cropping;
anamorphic compression or expansion;
panoramic or spherical mapping.

Two cameras at the same viewpoint with different sensor formats may therefore record different portions of the same underlying optical projection.

Focus, Depth of Field and Resolution

Camera View / Image also contains optical properties that should be distinguished from projection geometry.

Focus determines which object distance is sharply imaged.

Depth of field describes the range of object distances reproduced with acceptable sharpness.

Resolution describes the degree of fine spatial detail that the complete imaging system can distinguish.

These factors affect how much structural information becomes visible in a Camera Image but should not automatically be confused with changes in viewpoint or perspective geometry.

Exposure and Motion

A Camera Image represents not only space but also an interval of time.

Exposure duration can affect the visible appearance of moving objects:

a short exposure can freeze rapid motion;
a long exposure can record motion blur;
very long exposures can accumulate information too faint to be visible instantaneously;
high-speed cameras can divide very rapid events into sequences of separate frames;
time-lapse imaging can compress very slow changes into a visible moving sequence.

The camera can therefore reveal temporal structures inaccessible to ordinary human viewing, just as telescopes and microscopes reveal spatial structures inaccessible to ordinary vision.

Perspective Distortion and Lens Distortion

Several very different effects are often loosely described as photographic distortion. They should be distinguished.

Viewpoint or perspective effects result from camera position relative to the scene.
Projection effects result from the geometrical mapping used to place an angular field onto the image surface.
Camera-orientation effects include convergence caused by tilting the camera relative to sets of parallel scene lines.
Optical lens distortion results from departures of a real lens from its intended projection.

Common optical lens distortions include barrel, pincushion, moustache and decentring distortion.

A wide rectilinear image may also exhibit strong lateral stretching of objects near its margins even when the lens itself is geometrically well corrected. This results from mapping a very wide angular field onto a flat plane while retaining straight-line projection.

These phenomena should not all be attributed simply to a “wide-angle lens”.

Stereo and Multi-Camera Images

A conventional single camera records a principally monocular view. Additional cameras can provide multiple simultaneous viewpoints.

A stereo camera employs two separated imaging positions to produce corresponding left and right images. When delivered separately to the two eyes, their binocular disparity can contribute to the perception of stereoscopic depth.

Larger arrays of cameras can provide:

multi-view imagery;
photogrammetric reconstruction;
light-field imaging;
volumetric capture;
360° imaging;
machine depth estimation;
three-dimensional scene reconstruction.

The camera image is therefore not restricted to the traditional single-lens photograph.

Panoramic and 360° Camera Images

Panoramic cameras extend the recorded field beyond that of an ordinary framed photograph.

This can be achieved using:

wide-angle or fisheye lenses;
rotating cameras;
multiple synchronised cameras;
mirror-and-lens systems;
sequential image stitching;
computational spherical imaging.

A 360° camera can record an entire horizontal surround and, in many systems, almost the complete surrounding sphere.

Such imagery requires a projection or mapping system for storage and display, commonly cylindrical, spherical or equirectangular.

Processing Can Change the Camera Image

A modern Camera Image is frequently altered after its initial optical formation.

Processing may change:

brightness and contrast;
colour and white balance;
sharpness;
noise;
dynamic range;
local tone;
image geometry;
lens distortion;
perspective correction;
cropping;
resolution;
depth information;
image content itself.

Some computational cameras combine several exposures or several viewpoints into one output. The final displayed picture may therefore be a composite or computational image rather than the untouched result of one optical exposure.

Camera Image and Human Vision

A Camera Image should not be assumed to be identical to human visual experience.

A conventional photograph typically provides:

one fixed viewpoint;
one bounded image frame;
a particular projection geometry;
a fixed exposure interval;
a particular focus and depth of field;
a finite image resolution.

Human visual experience instead involves two moving eyes, a curved retinal surface, repeated eye and head movements, a wide changing visual field, binocular information, accommodation, varying acuity and continuous perceptual interpretation.

Camera Perspective and Visual Perspective therefore share important optical and geometrical principles but should not be treated as identical systems.

Explore Visual Perspective →

Camera Images in Computer and Robotic Vision

For a machine-vision system, a Camera Image may be an intermediate source of spatial data rather than a final picture intended for human viewing.

Computer and robotic vision systems can analyse camera images to estimate or identify:

objects;
edges and surfaces;
motion;
distance and depth;
position;
orientation;
scale;
three-dimensional structure;
obstacles;
routes and navigable space.

Stereo cameras, structured-light systems, depth cameras and multi-camera arrays can provide additional information from which three-dimensional spatial relationships are reconstructed.

In this context the Camera Image becomes part of a larger chain:

physical scene → camera image → computer analysis → spatial model → machine decision or action

Satellite and Remote-Sensing Camera Images

Cameras also form views from aircraft, drones, satellites and spacecraft, greatly extending the possible scale and position of the camera viewpoint.

Remote-sensing images may record visible light or other regions of the electromagnetic spectrum and may be used for:

mapping;
weather observation;
environmental monitoring;
agriculture;
geology;
oceanography;
urban analysis;
military reconnaissance;
planetary and astronomical exploration.

Such Camera Images may be geometrically corrected, mosaicked, colour-mapped and combined with coordinate data before they are interpreted or displayed.

The camera has therefore become both a visual instrument and a major spatial measurement and data-acquisition system.

Image Space is not Object Space

A Camera Image is a representation formed in image space. It should not be confused with the three-dimensional object space from which it originated.

Projection transforms spatial relationships. Three-dimensional depth is mapped into an image; surfaces may overlap; hidden regions disappear; apparent shape changes with orientation; and objects of different physical sizes and distances can produce similar image sizes.

Consequently, a single monocular Camera Image does not normally contain enough information to reconstruct one unique physical scene without additional assumptions or information.

This is part of the wider Correspondence or Equivalence Problem of perspective: different physical arrangements can potentially generate the same or very similar image.

A Camera Image is not a Neutral Copy of Reality

A Camera Image is sometimes described as if the camera simply copies whatever stands in front of it. In fact, every camera system selects and transforms information.

The recorded appearance depends upon choices or constraints involving:

where the camera is placed;
which direction it faces;
which projection is used;
which field of view is selected;
which part of the projection is framed;
what is in focus;
how long the exposure lasts;
what spatial resolution is recorded;
what wavelengths or signals are detected;
how the data is processed;
how the final image is displayed.

Different cameras can therefore produce substantially different but equally legitimate images of the same physical reality.

Camera View / Image across Perspective

Camera View / Image operates across many branches of perspective, including:

Camera Perspective;
Photographic Perspective;
Instrument Perspective;
Optical Perspective;
Cinema and Motion Perspective;
Stereoscopic Perspective;
Panoramic and Spherical Perspective;
Scientific Imaging;
Computer Vision;
Robotic Vision;
Photogrammetry;
Remote Sensing;
Digital and New Media Perspective;
Virtual and Augmented Reality.

The camera is therefore not simply a photographic appliance. It is one of the principal perspective instruments through which modern humans and machines view, capture, measure, analyse and communicate spatial reality.

Why Camera View / Image Matters

Understanding the distinction between a camera, its live view and its recorded image helps separate several concepts that are often conflated.

The camera is the imaging instrument.
The Camera View is the live or real-time perspective view formed through the instrument.
The Camera Image is the captured or recorded perspective product.
Camera Perspective concerns the spatial and optical relationships governing how that view or image is formed.
Image processing may subsequently transform the captured image.
Display presents the image in a new physical form.
Human or machine vision finally interprets or analyses it.

This distinction makes it possible to analyse the complete imaging process without confusing the spatial reality, the imaging method, the live view, the recorded image, the displayed representation and the final act of viewing.

Explore Theory

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

Explore Applications

Main Applications routes:
Applications of Perspective · Spatial Themes · Instruments of Perspective · Applications by Discipline · Perspective Studies Today

Major discipline hubs:
Art and Perspective · Architecture and the Built Environment · Photography · Cinema, Television and Visual Effects · Computer Graphics, Games and Extended Reality · Cartography, GIS and Spatial Mapping · Science, Engineering and Technical Imaging · Medical Imaging · Perspective in Professional Practice

AI-Generated Images →