Industrial Surveillance Cameras

What is an Industrial Surveillance Camera

An industrial surveillance camera is a fixed or steerable imaging device engineered to operate continuously in harsh environments, capturing video for security monitoring, process supervision, safety compliance, and evidential recording. Unlike a consumer webcam, it is specified against documented operational requirements and built to survive dust, water, vibration, temperature extremes, and in some cases explosive atmospheres for a service life measured in years rather than months.

Functionally a modern IP camera is a complete signal chain in one enclosure. Light passes through a lens onto an image sensor (today almost always a CMOS sensor), which converts photons to a digital pixel array. An image signal processor (ISP) handles exposure, white balance, wide dynamic range, and noise reduction. A video encoder compresses the raw frames with H.264 or H.265, an analytics engine may run object detection or line-crossing rules on the edge, and a network interface streams the result as RTSP or ONVIF over Ethernet or fibre. This is fundamentally different from legacy analog CCTV, where the camera produced only a composite video signal sent over coaxial cable to a digital video recorder (DVR) that did all the encoding.

The history of the technology runs from analog to digital to intelligent. Closed-circuit television emerged in the 1940s and 1950s for monitoring of rocket launches and industrial processes. Charge-coupled device (CCD) sensors in the 1980s replaced vacuum tubes and made compact solid-state cameras practical. The first commercial network camera, the Axis NetEye 200, appeared in 1996, beginning the shift from analog CCTV to IP video. Through the 2000s, megapixel sensors, H.264 compression, and Power over Ethernet made IP the default for new installations. Since the late 2010s, on-camera deep-learning analytics, the IEC 62676 standard series, and cybersecurity hardening have defined the current generation.

In industry, the camera rarely works alone. It is one node in a video surveillance system that includes the network switch (often Power over Ethernet), the recording layer (an NVR or a server running a VMS), storage, and the analytics or alarm management software. The IEC 62676 series formalises this view: Part 1 covers system requirements, Part 2 video transmission protocols, Part 3 analog and digital interfaces, Part 4 application guidelines including the DORI pixel-density criteria, and Part 5 standardised image quality and data specifications for camera devices. Treating the camera as a system component, not a standalone gadget, is the first discipline of correct selection.

Four engineering dimensions ultimately govern whether a camera meets its purpose: image quality at the target distance (resolution, low-light performance, dynamic range), environmental survivability (IP, IK, temperature, corrosion, hazardous-area rating), interoperability (ONVIF profile, codec, VMS compatibility), and total cost of ownership including power, cabling, storage, and cybersecurity maintenance over the install lifetime. A camera that wins on megapixels but fails on dynamic range, sealing, or firmware support is a poor industrial choice.

Camera Form Factors and Types

Camera form factor is the first practical selection axis because it bundles together the housing, mounting, lens flexibility, and tamper resistance that suit a given location. The wrong form factor leads to blind spots, vandalism, or impractical mounting, regardless of how good the sensor is. The table below compares the mainstream industrial form factors.

Bullet cameras are long cylindrical units with a fixed or motorised varifocal lens and an integral sun shroud. Their shape makes them easy to aim along a defined line of sight and visible as a deterrent, which suits perimeters, fence lines, loading bays, and car parks. The exposed form is less vandal-resistant than a dome, so bullets are usually mounted out of reach.

Dome and turret cameras place the optics behind a hemispherical bubble or a low-profile turret eyeball. The enclosed shape resists tampering and makes the aiming direction hard to read, which is why vandal-resistant domes (IK10) dominate public-facing indoor and ceiling installations. Turrets reduce the internal IR reflection and condensation problems that can affect bubble domes.

PTZ (pan-tilt-zoom) cameras add motorised movement, typically 360 degrees continuous pan, a wide tilt range, and a high optical zoom of 20x to 40x or more. They cover large areas from a single point and can track a moving target or follow an analytics or alarm trigger, but they only see where they are pointed, so they are best paired with fixed cameras that provide continuous wide coverage. PTZ models draw more power and contain wear parts (motors, slip rings) that affect serviceability.

Fisheye and multi-sensor panoramic cameras deliver a 180 or 360 degree overview from one unit, with software dewarping to flatten the hemispherical image. They reduce camera count for open areas but trade per-target pixel density for coverage. Box and modular cameras separate the sensor body from the lens and housing, allowing specialty optics or engineered enclosures, and are the basis of most explosion-proof and high-temperature builds. Thermal cameras, covered in detail in Chapter 3, form their own class for darkness and long-range detection.

Imaging and Encoding Technology

Image quality is set by the chain of sensor, lens, image processing, and video compression, and each link has its own engineering trade-offs. Two cameras with the same megapixel count can deliver completely different results once low light, motion, and high-contrast scenes are introduced. The table below compares the core imaging technologies an industrial buyer should understand.

The image sensor is the heart of the camera. Industrial IP cameras use CMOS sensors in formats from about 1/2.8 inch up to 1/1.8 inch and larger. A bigger sensor with a larger pixel pitch collects more light per pixel, which is why a 2 MP camera on a large sensor can outperform an 8 MP camera on a small one in dim conditions. Marketing emphasises megapixels, but for night-time and low-contrast scenes the sensor format and pixel pitch matter more than raw resolution.

Resolution and frame rate set both detail and data load. Common industrial resolutions are 2 MP (1920x1080), 4 MP (2560x1440), 5 MP (2592x1944), and 8 MP, which is 4K UHD at 3840x2160. Higher resolution extends the DORI range (Chapter 5) but multiplies bitrate and storage. Frame rate of 25 to 30 fps is standard; 50 to 60 fps is reserved for fast motion such as gaming floors or production lines.

Wide dynamic range (WDR) lets the camera hold detail in a scene that contains both bright and dark areas, such as a doorway against sunlight. True WDR captures multiple exposures and combines them, reaching 120 to 140 dB on high-end sensors, while digital or DWDR only stretches tone curves on a single exposure and is far weaker. Low-light performance is quoted as minimum illumination in lux; starlight-class sensors produce usable colour images down to around 0.001 to 0.01 lux, and most cameras switch to monochrome with built-in infrared LEDs at full darkness.

Video compression determines how much network bandwidth and storage the camera consumes. H.264 (AVC) remains widely supported, but H.265 (HEVC) achieves the same image quality at roughly 30 to 50 percent lower bitrate. Vendor smart-codec techniques (Axis Zipstream, Hanwha WiseStream, generic smart codec) add region-of-interest and dynamic GOP control that can cut bandwidth a further 30 to 70 percent on static scenes. The trade-off is that aggressive smart-codec settings can soften detail on the very motion you want to capture, so settings should be validated against the operational requirement.

Thermal imaging is a separate technology rather than a low-light option. An uncooled microbolometer detects long-wave infrared radiation (roughly 8 to 14 micrometres) emitted as heat, forming an image with no visible light at all. Thermal cameras see through total darkness, light fog, and smoke and excel at detecting a warm intruder against a cool background over long perimeters, which is why series such as the Axis Q19 and Q21 thermal families (for example the Q1961-TE and Q1971-E) are used for perimeter protection. They cannot resolve faces or licence plates, so they are paired with visible cameras for identification.

Environmental and Hazardous-Area Protection

For industrial deployment, environmental protection is as decisive as image quality. A camera that resolves perfectly but admits dust, fails at minus 30 degrees Celsius, or cannot be installed in a Zone 1 area is useless on site. Three families of standards govern this: IEC 60529 for ingress protection, IEC 62262 for impact resistance, and IEC 60079 for explosive atmospheres. The table below summarises the protection ratings an industrial specifier reads first.

Ingress protection (IP) uses a two-digit code from IEC 60529. The first digit rates solids: 5 is dust-protected and 6 is fully dust-tight. The second rates water: 6 means powerful water jets, 7 means temporary immersion to 1 m for 30 minutes, and 8 means continuous immersion under conditions agreed with the maker. Outdoor industrial cameras are typically IP66 or IP67. IP66 handles driving rain and pressure-washing; IP67 adds immersion tolerance for flood-prone or washdown areas such as food processing.

Impact protection (IK) from IEC 62262 runs IK00 to IK10. IK08 withstands a 5 joule strike and IK10 withstands 20 joules, roughly a 5 kg mass dropped from 40 cm. IK is independent of IP: a camera can be fully weatherproof yet not vandal-resistant. Public-facing and reachable cameras should carry both an IP66 seal and an IK10 housing. Some armoured domes are now rated to the higher IK10+ or IK11 levels offered by individual makers, though IK10 is the top tier of the formal IEC scale.

Operating temperature and corrosion matter for outdoor and process environments. Industrial cameras commonly specify a wide operating range such as minus 40 to plus 60 degrees Celsius, with internal heaters and demisting for cold starts and condensation. Coastal, chemical, and washdown sites need corrosion-resistant construction; 316L stainless steel housings and the North American NEMA 4X rating address salt spray and aggressive cleaning agents that would pit ordinary die-cast aluminium.

Hazardous-area (explosion-proof) protection is mandatory wherever flammable gas, vapour, or combustible dust can be present: oil and gas, refineries, petrochemicals, fuel depots, offshore platforms, paint shops, and grain handling. The IEC 60079 series classifies the location by zone: Zone 0, 1, 2 for gas (continuous, likely, unlikely) and Zone 20, 21, 22 for dust. The dominant protection method for cameras is the flameproof enclosure, marked Ex db (formerly Ex d), a heavy 316L stainless steel housing engineered so that an explosion ignited inside cannot propagate to the surrounding atmosphere. The certificate also states the gas group (IIA, IIB, IIC, where IIC including hydrogen is the most demanding) and the temperature class T1 to T6, which caps the maximum surface temperature. ATEX is the mandatory EU framework under directive 2014/34/EU; IECEx is the equivalent international certification scheme; North America uses the Class/Division system (for example Class I Division 1). A hazardous-area camera must carry the certificate that matches the specific zone, gas group, and temperature class of the installation.

Key Specification Parameters

Reading a camera data sheet is a fundamental procurement skill. A single spec sheet may list dozens of parameters, but a manageable set drives the selection decision: pixel density against the operational requirement, resolution, sensor and low-light figures, dynamic range, codec and bitrate, lens and field of view, protection ratings, power, and interoperability. Each is decoded below.

DORI pixel density is the single most important and most misread spec, defined in IEC 62676-4. It states how many pixels per metre the camera places across a target at a given distance, and maps that to four operational tasks. Detect needs 25 px/m (confirm something is present), Observe needs 63 px/m (follow activity), Recognise needs 125 px/m (tell whether a person is already known), and Identify needs 250 px/m (establish the identity of an unknown person to evidential standard). The note that follows the 2025 revision of the standard expands these into a finer scale, but the four DORI levels remain the practical reference. Because pixel density falls with distance and widens with the lens, a 4 MP camera may identify at 5 m but only detect at 50 m on the same lens. Always compute DORI at the actual target, using the manufacturer DORI chart, not the headline megapixel figure.

Resolution and sensor together determine detail and low-light capability. Read both the megapixel figure (2, 4, 5, 8 MP) and the sensor format (1/2.8 to 1/1.8 inch) and pixel pitch. For night work, a larger sensor with fewer, larger pixels often beats a higher-resolution small sensor. Minimum illumination in lux and wide dynamic range in decibels quantify low-light and high-contrast performance; insist on true WDR (120 dB or more) for scenes with strong backlight.

Lens and field of view set coverage and DORI range. A fixed lens gives one field of view; a motorised varifocal lens covers a range, for example 2.8 to 12 mm, allowing the same model to suit different mounting distances. A wider lens covers more area but reduces pixel density per metre, so lens choice is inseparable from the DORI calculation.

Codec and bitrate define network and storage load. Confirm H.265 support and the smart-codec option, and read the rated bitrate at the intended resolution and frame rate. A 4 MP stream at 25 fps on H.265 typically runs 2 to 6 Mbit/s; design the network switch, uplink, and storage for the aggregate peak across all cameras, not the average.

Power is usually Power over Ethernet, which carries data and power on one cable up to 100 m. IEEE 802.3af (PoE) supplies up to 15.4 W and suits fixed cameras drawing 4 to 12 W; IEEE 802.3at (PoE+) supplies up to 30 W for heaters and IR; IEEE 802.3bt (PoE++) supplies up to 60 or 90 W for high-speed PTZ with heaters and wipers. Explosion-proof and large PTZ units often need separate 24 VAC or 24 VDC supplies. Size the PoE switch budget for the sum of all ports plus margin.

Interoperability is governed by ONVIF profiles: Profile S for basic H.264 streaming and PTZ, Profile T for H.265, multiple streams, and analytics metadata, Profile G for edge recording and retrieval, and Profile M for metadata and events. Match profiles on both the camera and the VMS or NVR. Cybersecurity features (signed firmware, TLS, IEEE 802.1X port authentication) and supply-chain compliance such as NDAA are now first-class specs on industrial and government tenders, not optional extras.

Selection Decision Factors

To turn the preceding five chapters into a specific model, work through the decision sequence below. Most selection mistakes come not from a single wrong answer but from skipping the operational-requirement step and jumping straight to a product. These eight steps form a reusable RFQ template.

1. Define the operational requirement first: for each viewpoint, state whether you need to Detect, Observe, Recognise, or Identify, and at what distance. This sets the DORI pixel density (25, 63, 125, or 250 px/m) that every later choice must satisfy.
2. Derive resolution and lens from DORI: use the manufacturer DORI chart to pick the megapixel count and lens (fixed or motorised varifocal) that deliver the required pixel density at the target distance, then check the resulting field of view covers the scene.
3. Choose the form factor: bullet for directional outdoor lines, dome or turret (IK10) for reachable public areas, PTZ for active tracking over large yards, fisheye for whole-room overview, thermal for darkness and long-range detection. Pair fixed coverage with PTZ where both are needed.
4. Set environmental and protection ratings: IP66 or IP67 outdoors, IK10 where vandalism is a risk, operating temperature for the site climate, and 316L stainless or NEMA 4X for corrosive or washdown areas.
5. Confirm hazardous-area certification if required: classify the zone per IEC 60079 (Zone 1 or 2, Zone 21 or 22), then specify an ATEX or IECEx flameproof (Ex db) camera with the correct gas group (IIA/IIB/IIC) and temperature class (T1 to T6), or the matching Class/Division rating in North America.
6. Specify imaging quality: true WDR (120 dB or more) for high-contrast scenes, minimum illumination and IR or starlight for night, codec (H.265 plus smart codec) and frame rate for the activity level.
7. Lock interoperability and power: match ONVIF profile (S, T, G, M) to the VMS or NVR, confirm codec support, and size PoE (802.3af / at / bt) or auxiliary 24 V supply, including the switch power budget and cable runs.
8. Evaluate total cost of ownership and cybersecurity: add cabling, switch, storage (bitrate times retention), licensing, and firmware-maintenance cost over the install life, and require signed firmware, TLS, 802.1X, and supply-chain compliance such as NDAA.

One dimension that is easy to underrate is manufacturer serviceability and cybersecurity support: long-term firmware and security patching, local spare-part availability for PTZ wear parts, documented vulnerability disclosure, and proven VMS integration. On a 10-year industrial install these determine uptime and exposure far more than the original optics. Axis Communications, Hanwha Vision (Wisenet), Bosch, Pelco, and Hikvision lead the general industrial market with broad ONVIF-conformant ranges and edge analytics; Axis Q19/Q21, FLIR, and Teledyne FLIR lead thermal perimeter detection; and ATEX/IECEx 316L models from Pelco, ExLoc, Defendex, and A2S cover hazardous areas. Validate a sample unit against your VMS before any bulk order.