What is an area scan camera?

What an area scan camera is

An area scan camera is an industrial camera that captures a full two-dimensional image in a single exposure. The sensor is a rectangular grid of pixels, rows and columns, so one trigger produces one complete frame: height × width, just like a photograph, but built for machines that run all day.

That simple idea is why area scan dominates most Computer Vision stations. If your part fits in a field of view, sits still (or moves slowly enough) during exposure, and the defect or feature is visible in a 2D image, an area scan camera is usually the right starting point. You mount a lens, light the scene, trigger on a PLC pulse or encoder, and send frames to a PC, smart system, or vision controller.

In product catalogues you will also see labels such as matrix camera, 2D industrial camera, or simply Computer Vision camera. In Sedeco’s taxonomy those sit under Cameras, with area scan as the default family for discrete-part inspection.

How an area scan camera works

Think of the optical path as four linked decisions: scene → lens → sensor → interface.

1. The scene and the trigger

A part arrives in the station. A sensor, PLC, robot, or encoder tells the camera “now.” During the exposure time, photons from the illuminated scene hit the sensor. If the part moves too far during that window, the image blurs, which is why lighting, shutter type, and exposure time matter as much as megapixels.

2. The lens forms the image

The lens maps object size and working distance onto the sensor. Focal length, aperture, distortion, and resolution class decide whether a 0.2 mm scratch is even resolvable. Camera choice without lens choice is incomplete, see our lenses overview when you size field of view.

3. The sensor samples the frame

An area scan sensor is a 2D array. Each pixel integrates light for the exposure. Mono sensors measure intensity; color sensors typically use a Bayer filter and interpolate RGB. Pixel size (often a few micrometers) and sensor format (for example 1/2.9", 1/1.8", or larger) affect sensitivity, depth of field via optics, and how much of the optical circle you use.

4. The interface delivers the image

After readout, the camera streams the frame over GigE Vision, USB3 Vision, Camera Link, CoaXPress, or another industrial interface. GenICam-style control lets software set exposure, gain, trigger mode, and ROI without proprietary lock-in on most modern cameras.

Area scan vs line scan, smart, and 3D

Teams often ask for “an industrial camera” when they actually need a family decision. Use this comparison as a filter before you open a datasheet.

Area scan vs line scan: if the product is a discrete object that stops or moves slowly in a window, area scan is simpler. If material streams continuously and you need microscopic defects along meters of product, line scan is usually the tool. Many factories use both, area scan on assembly stations, line scan on webs.

Area scan vs smart camera: a “plain” area scan camera delivers images to a PC or controller. A smart camera embeds tools on the device. Architecture comes first, see also our smart camera guide and smart camera vs PC-based vision, then you pick the sensor family.

Area scan vs 3D: area scan answers “what does it look like in 2D?” 3D answers “what is its shape and pose?” Surface print defects often need area scan; grasping randomly piled parts often needs 3D.

Where area scan cameras are used

If you are buying Computer Vision for the first time, these patterns cover most successful area-scan projects:

• Presence and assembly checks, Is the clip seated? Are all screws present? Did the label land in the correct zone?
• Surface and cosmetic inspection, Scratches, stains, incomplete coatings, flash on molded parts, print quality on packaging.
• Dimensional gauging (2D), Width, diameter, gap, position relative to a datum, when tolerance fits a calibrated 2D view.
• Identification, 1D/2D codes and OCR on cartons, PCBs, vials, and nameplates (sometimes better as a dedicated code reader, but area scan + software remains common).
• Robot guidance (2D), Find a part’s XY and rotation on a belt or fixture so a robot can pick or assemble.
• Electronics and precision assembly, Connector pins, solder joints (with the right optics/light), component orientation, and PCB features.

Industries that lean heavily on area scan include packaging and FMCG, automotive component cells, pharma secondary packaging, electronics, and general machine building. The common thread is a bounded FOV and a decision that can be made from a still (or motion-frozen) 2D image.

Specs that actually matter when you buy

Datasheets are dense. Prioritize the list below in order. Megapixels alone rarely decide success.

Resolution and pixel size

Resolution is the number of pixels across the image (for example 1920×1200). Rough rule: to reliably measure or detect a feature, plan on several pixels across the smallest critical detail, often more if lighting or contrast is weak. Pixel size affects light sensitivity and how optics map millimeters on the part to pixels on the sensor. Higher resolution is not automatically better: it costs bandwidth, storage, and processing time, and may force a slower frame rate.

Sensor format and lens mount

Sensor size must match the lens image circle. C-mount is common on industrial area scan cameras; larger sensors may need other mounts. Undersizing the lens for the sensor causes vignetting; oversizing can waste optics budget. Treat camera + lens as one optical design.

Frame rate vs resolution

Maximum frame rate falls as resolution and bit depth rise, and as the interface saturates. If takt time needs 30 good images per second at full resolution, verify that number under your pixel format, not only the marketing “max fps” at a tiny ROI. Using a region of interest (ROI) is a legitimate way to go faster when you only care about part of the sensor.

Mono vs color

Mono cameras generally deliver higher effective resolution and sensitivity for the same sensor class, prefer mono for metrology, code reading with controlled light, and most defect detection. Choose color when the defect or class is defined by color (wrong wire color, print color, material tint) or when operators need a natural image. Bayer color interpolates; do not assume color equals “more information” for every task.

Interface: GigE, USB3, and friends

GigE Vision is the workhorse for machine builders: long cables (typically up to ~100 m with appropriate networking), easy multi-camera topologies, and industrial switches. USB3 Vision is excellent for short runs, lab setups, and compact machines near the PC, simpler cabling, but length and hub behavior need care. Camera Link and CoaXPress serve higher bandwidth or specialized lines. Pick the interface for cable plant and PC architecture, not brand preference.

Global vs rolling shutter

Global shutter exposes all pixels at once, preferred when parts move during exposure. Rolling shutter reads lines sequentially and can skew fast motion. Many modern CMOS area scan cameras offer global shutter; confirm it if conveyors or robots move during the flash.

Lighting dependency

The best camera cannot fix bad contrast. Ring, bar, backlight, dome, and strobe geometries decide whether a scratch appears. Budget time for lighting trials, start from our lighting page when you design the station.

Building a working area-scan setup

A reliable station is a system, not a SKU:

1. Camera, sensor, shutter, interface matched to takt and FOV.
2. Lens, focal length and resolution class for the feature size.
3. Lighting, continuous or strobed, often the make-or-break element.
4. Mechanics, stable mount, vibration control, cable strain relief.
5. Triggering, PLC, photoelectric, encoder; consistent part presentation.
6. Compute & software, PC, vision box, or smart path; GenICam drivers; recipe management.

Common first-system mistakes

• Buying megapixels before measuring the required resolution on real samples.
• Ignoring motion blur, then “fixing” it with gain that adds noise.
• Using rolling shutter on a fast conveyor without strobe strategy.
• Underspecifying lighting and expecting software to invent contrast.
• Choosing USB3 for a 40 m cable run (use GigE or redesign the layout).
• Skipping sample variation, shiny, dark, and dirty parts all appear on Monday morning.

How to read an area scan product page

On Sedeco product pages for area scan families you will typically see options such as sensor model, resolution, and max. frame rate, plus interface cues in the SKU or specifications (GigE vs USB, mono vs color). Read them as answers to the eight questions above:

• Sensor model, the chip; implies pixel size, shutter behavior, and optical format.
• Resolution, pixels available for your FOV; combine with optics to get µm/pixel.
• Max. frame rate, ceiling under stated conditions; confirm for your pixel format and ROI.
• Data interface, how frames reach the PC or controller.
• Mono/color, intensity vs Bayer color; match to the inspection logic.

When you evaluate “area scan camera,” that phrase can link here so you can understand the category before comparing models in the camera catalogue.

Worked example: sizing resolution

Suppose you inspect a molded cover that is 80 mm wide in the field of view. The critical defect is a 0.4 mm flash on the edge. You want at least 5 pixels across that flash so detection is stable when contrast varies.

Required spatial sampling: 0.4 mm / 5 pixels = 0.08 mm per pixel. Across 80 mm you need 80 / 0.08 = 1000 pixels of width, so a 1.3 MP (for example 1280×1024) sensor can work if the FOV is tight and optics are sharp. If the FOV must be 160 mm to tolerate part placement, you need about double the horizontal pixels, or a different optical layout (two cameras, or accept coarser sampling).

This is why buying megapixels by habit is not a strategy. Start from feature size, FOV, and pixels-on-feature. Then confirm with real samples under real lighting. Sedeco often runs a short feasibility pass for exactly this reason, see the feasibility checklist.

Industry snapshots

Area scan shows up differently by sector. The sensor family stays the same; the lighting and acceptance criteria change.

Packaging and FMCG

Label presence, date-code OCR, seal integrity cues, fill-level windows, and carton assembly checks. Parts move on conveyors; strobe lighting and global shutter keep edges crisp. Color may matter for brand artwork; mono often wins for codes and geometry.

Automotive components

Clip seating, thread presence, gasket position, pore or scratch checks on machined faces. Stations are often robot-fed with tight takt. Calibration and stable mounts matter as much as megapixels when you gauge in 2D.

Electronics

Connector pin count and coplanarity cues (sometimes better with 3D), polarity marks, PCB feature checks, and solder cues with coaxial or dome light. High resolution and good lenses are common; heat and ESD practices affect housing choice.

Pharma secondary packaging

Leaflet presence, carton codes, aggregation support, and print verification. Validation and recipe control dominate the project, the camera must be reproducible, not merely impressive on a datasheet alone.

Machine builders and OEMs

You standardize on a camera family (often GigE area scan) across machines shipped worldwide. GenICam compatibility, PoE options, and long-term availability weigh as heavily as peak fps. Document the optical recipe so field service can repeat it.

Integrating with PLC, robot, and IT

An area scan camera rarely stands alone. Typical industrial loops look like this:

• Trigger in, photoelectric or PLC output fires the camera (or light controller + camera).
• Acquire, exposure under strobe; image to PC or smart device.
• Decide, vision software returns pass/fail, measurements, or pose.
• Act, PLC rejects, robot offsets, or MES logs a serial.

Latency budgets matter: image transfer + algorithm + handshake must fit inside takt with margin. Multi-camera GigE systems need switch quality and NIC tuning; USB3 systems need short, reliable cabling and attention to host controllers.

If you are unsure whether processing should live on a smart camera or a PC, settle that architecture question before freezing the area-scan SKU, again via the smart vs PC decision page.

Cost, quality, and what good enough means

Teams sometimes under-buy optics and lighting to save on the camera, then spend months fighting false rejects. The inverse also happens: an expensive sensor with a kit lens and ambient light. Image quality is a chain; the weakest link sets the false-reject rate.

Practical budgeting for a first station often allocates meaningful spend to lighting and mechanics, mounts, enclosures, cable management, not only to the camera body. A stable 5 MP GigE mono camera with the right ring light and lens routinely outperforms a much higher-megapixel camera in uncontrolled light.

Define acceptance with samples: golden parts, borderline fails, and dirty or reflective variants. If you cannot show the defect to a human under the proposed lighting, software will struggle too.

Short glossary

FOV

Field of view, the width/height of the world seen by the camera.

WD

Working distance, lens front to object (definition varies slightly by vendor).

ROI

Region of interest, readout of only part of the sensor to gain speed.

GenICam

Standard control model for many industrial cameras.

Bayer

Color filter pattern on many color CMOS sensors.

Global shutter

All pixels expose together, preferred for moving parts.

Strobe

Short bright pulse of light synced to exposure, freezes motion.

FAQ: area scan cameras

How this guide fits your selection

Use this guide when you are still deciding whether area scan is the right family, before you compare model options. It answers the practical questions: Is this the right approach? Why GigE or USB3? Mono or color? What else does the station need?

Sedeco will roll out the same depth for line scan, 3D, smart cameras, code readers, lenses and lighting, each under Knowledge, and each linked from the matching product hub such as Cameras.

Next steps

You now have the category model: area scan captures a full 2D frame for discrete inspection, measurement, ID, and guidance, provided optics and lighting are designed with the camera.

Related knowledge hub: Kennis / Knowledge. Continue with line scan, 3D cameras, smart cameras, code readers, lenses and lighting.