Lighting for security cameras must support the image sensor as well as the person viewing the scene. A space that appears acceptably lit to the human eye may still produce silhouettes, glare, motion blur or loss of detail on camera, so security lighting has to be designed around the visual task the system is expected to perform.
Human vision adapts continuously to changes in brightness, colour and contrast. A person can look from a bright entrance towards a darker interior, shift attention between foreground and background, and use context to interpret an incomplete view. A camera has a more limited response. Its exposure, lens, sensor, processing and compression settings determine what is recorded, and strong differences within the scene can cause important detail to disappear.
The position of light also matters differently. General workplace lighting is often assessed on a horizontal plane because it supports movement or work on desks and floors. Security cameras frequently need useful vertical illumination on faces, clothing, vehicle sides or identification features. A well-lit floor does not guarantee that a face approaching a doorway will be visible to a camera mounted above it.
The design objective should therefore begin with the required security outcome. Lighting that is adequate for detecting movement may not be adequate for recognising a familiar person, while recognition may not provide enough detail for confident identification. The camera view, expected subject position and intended use of the recorded image should be considered together.
Detection means establishing that a person, vehicle or object is present. The image may show movement, size and general direction without revealing meaningful detail. This is suitable for alerting an operator or triggering a response, but it is not the same as being able to say who or what is present.
Recognition means distinguishing whether a person or object is familiar or whether it matches a known category. A camera may show enough information to recognise a regular employee, a particular vehicle type or a repeated pattern of activity. Recognition requires greater image clarity than simple detection and is more vulnerable to poor contrast, backlighting and motion.
Identification is the most demanding purpose. It depends on sufficient detail, suitable viewing angle, controlled glare and a face or other identifying feature being presented clearly to the camera. A system described as providing identification should be evaluated at the actual point where identification is expected to occur, rather than inferred from a general view of the site.
These purposes should not be treated as interchangeable. A wide-angle camera covering a large area may be effective for detection but unsuitable for identification at the edge of the image. Lighting should be matched to the relevant field of view and the distance at which the security decision must be made.
Vertical illuminance describes light falling on an upright surface, such as a face, door, wall or vehicle side. For security imaging, it is often more relevant than light measured on the ground. A camera positioned above eye level may already see faces at a steep angle, so poor vertical lighting can leave eye sockets, facial features and clothing details in shadow.
Uniformity affects the camera's ability to maintain a useful exposure across the scene. Bright pools of light separated by darker zones can force the camera to choose between overexposing the bright area and losing detail in the shadows. Automatic exposure may change abruptly as a person moves between zones, producing a sequence in which the subject is visible in one frame and indistinct in the next.
A more even distribution of light generally gives the camera a more stable image, but uniformity should not be pursued without considering direction and contrast. Completely flat lighting can reduce modelling of faces and objects, while strongly directional lighting can create deep shadows. The useful arrangement is one that reveals form without producing extreme brightness differences.
Camera settings can compensate for some variation, but they cannot recover information that was never captured. Wide dynamic range processing, gain control and image enhancement may improve a difficult scene, yet each can introduce noise, blur, artefacts or reduced colour reliability. Lighting should therefore reduce the burden placed on image processing rather than assuming software will correct every defect.
Entrances are common failure points because they often combine bright exterior daylight with a comparatively dark interior. A camera facing a glazed door may expose for the bright background and render an entering person as a silhouette. The face can then remain unreadable even though the overall image appears bright.
The problem can also occur in reverse at night. A brightly lit interior viewed from a darker exterior can dominate the exposure, while a person standing outside remains underexposed. Headlights, illuminated signs and reflections from glazing can create similar effects by placing intense sources close to the subject within the camera image.
Lighting at entrances should therefore support the side of the person facing the camera and should be coordinated with camera position. Moving the camera, changing its angle or providing light from a more suitable direction can sometimes resolve a problem that additional general lighting would not. The aim is to avoid placing the subject between the camera and the brightest part of the scene.
Physical layout also influences the result. Canopies, vestibules, recessed doors and tinted glazing can moderate transitions, while glossy finishes may create reflections that obscure facial detail. A survey should examine the entrance under the conditions in which the camera is expected to operate, including strong daylight, dusk and night-time use.
Colour information can be important when security staff or investigators need to describe clothing, vehicles, packages or safety markings. Lighting with poor colour rendering can distort the apparent colour of an object, while mixed sources can produce inconsistent colour across the same scene. Camera white-balance settings may reduce some variation but cannot make unreliable spectral information fully accurate.
At low light levels, some cameras change from colour imaging to monochrome operation. This can improve sensitivity and reduce image noise, but it removes colour as an identifying characteristic. A dark red garment and a dark blue garment may appear similar in monochrome, and vehicle colours may no longer be described reliably.
Infrared illumination allows compatible cameras to record scenes that appear dark to the human eye. It can support detection and, in suitable conditions, recognition or identification. However, infrared images do not preserve normal colour appearance, and reflective materials, number plates, rain, dust, insects and nearby surfaces can create bright returns that reduce useful detail.
The relationship between infrared sources and the camera is therefore important. Illumination placed too close to the lens can encourage backscatter or reflections, while an object close to the camera may become overexposed. The night-time operating mode should be tested with realistic subjects and movement rather than judged only from a static empty scene.
A camera should not look directly into a luminaire, bright sign or low sun position where glare can reduce contrast across the image. Veiling glare may wash out fine detail, while flare and internal reflections can create bright shapes that obscure a subject. Even when the source is outside the recorded frame, light entering the lens at an oblique angle can still affect image quality.
Luminaire placement should illuminate the subject area without creating a bright source in the camera's principal view. Side lighting can reveal shape and texture, but excessively steep or rearward lighting can produce harsh shadows or silhouettes. The best arrangement depends on the camera angle, subject direction and surfaces within the scene.
Reflections also require attention. Polished floors, glass doors, wet roads, vehicle bodywork and metallic cladding can redirect light into the lens. In the UAE, dust accumulation and heat can further affect luminaires, enclosures and camera housings, so the maintained condition should be considered rather than only the appearance immediately after installation.
General outdoor lighting, obtrusive light and sky glow are addressed separately on the external and car park lighting page. For security and CCTV, the central question is whether the camera receives a stable, usable image for its intended purpose without glare, excessive contrast or loss of identifying detail.
In the Emirate of Abu Dhabi, workplace lighting is a mandatory matter. The Abu Dhabi Public Health Centre states that Codes of Practice under the ADOSH-SF framework are mandatory to all entities regardless of risk classification, and ADOSH-SF Code of Practice 8.0, General Workplace Amenities, Version 4.0, effective 15 July 2024, applies to all employers within the Emirate of Abu Dhabi. Its Section 3.8 sets minimum illumination intensities of 30, 10 and 5 foot-candles according to area type, which the Code itself converts at approximately 10.764 lux to the foot-candle. ADPHC's Technical Guideline on ergonomics in an office environment adds recommendations for office and screen work and is classified as non-mandatory. At federal level, MOHRE Administrative Decision No. 19 of 2023, Article 4(c), places lighting duties on employers within MOHRE's jurisdiction, including a requirement for emergency lighting when the primary system fails. EN 12464-1 and ISO/CIE 8995-1 are consensus standards: widely used and often specified contractually, but not statutory in the UAE unless an instrument adopts them. A requirement binding only in the Emirate of Abu Dhabi is not UAE law and must not be described as such.
ADOSH-SF Code of Practice 8.0 — General Workplace Amenities, V4.0 (15 July 2024), Section 3.8; MOHRE Administrative Decision No. 19 of 2023, Article 4(c)
Brightness alone does not determine image quality. Light may fall mainly on the ground rather than on faces, or strong bright and dark zones may exceed the camera's useful dynamic range. Glare, backlighting, motion and poor spectral quality can also reduce detail even when the scene appears bright.
Not usually. Horizontal measurement can describe light on the ground, but security cameras often depend on vertical illumination on faces, doors and vehicle sides. Measurements should relate to the surface and position that the camera must record.
Recognition means being able to decide whether a person or object is familiar or belongs to a known category. Identification requires enough detail to establish identity with greater confidence. The second purpose generally demands a more controlled camera view, lighting direction and subject position.
Infrared operation normally produces monochrome images or supports a camera operating without reliable visible-colour information. It can improve night-time visibility, but clothing and vehicle colours cannot be interpreted in the same way as under suitable visible light.
Image processing can improve some difficult scenes, but it has limits. It may increase noise, blur or artefacts, and it cannot recover detail that was not captured because a face was silhouetted, overexposed or hidden in shadow.
Daylight, dusk, darkness, headlights, signs and reflections can change the balance of a scene. A system that performs well under one condition may fail under another, so testing should reflect the full range of conditions in which detection, recognition or identification is expected.