Glossary — CameraCAD

Optics

Field of view (FOV): The angle a camera sees, usually quoted horizontally. Set by the focal length and the sensor's width together. Lens and field of view.
Focal length: The distance in millimetres from the lens's optical centre to the sensor when focused at infinity. Longer means narrower and further.
Fixed lens: A lens with one focal length. Cheap and robust; the view is chosen when the camera is bought.
Varifocal / motorised varifocal: A lens whose focal length is adjusted at install, by hand or remotely. Datasheets often quote its range at the long end; plan with the setting you will actually use.
Rectilinear projection: The ordinary lens model, in which straight lines stay straight and image position grows with the tangent of the angle. Breaks down beyond about 120°.
Fisheye (equidistant) projection: A lens in which image position grows in step with the angle, allowing 180° or more in one image at the cost of curved lines. The physics of a camera cone.
Sensor format: Sensor size named in the old tube convention, such as 1/2.8″. The number is not the real width: a 1/2.8″ sensor is about 5.4 mm across.
F-number: Focal length divided by the aperture's diameter. A lower number lets in more light: F1.0 gathers four times what F2.0 does.
Corridor format: Running the image rotated 90°, 9:16 instead of 16:9, so a tall narrow scene uses the whole sensor. Corridor format.

Pixel density (px/m): Horizontal pixels per metre of scene at the target's distance. The single number DORI is written in. Resolution versus detail.
DORI: Detect, observe, recognise, identify: the four operator tasks in EN 62676-4, at 25, 62.5, 125 and 250 px/m. DORI, explained.
Johnson criteria: A 1958 measure of how many line pairs across a target an observer needs to detect, recognise or identify it. Used for thermal and long-range cameras, and ten times more lenient than DORI. The Johnson criteria.
Line pair: One dark and one light bar; at least two pixels. The unit of resolution in the Johnson criteria.
Critical dimension: The target size a Johnson range is computed on, commonly 0.75 m for a standing person and 2.3 m for a vehicle.
Resolution: The sensor's pixel count, often given in megapixels. Sets the ceiling on pixel density; lens, light and compression decide how close you get to it.
Bitrate: The data rate of the compressed stream, in megabits per second. Too low and fine detail is the first thing lost.

Lux: Illuminance, lumens per square metre. Daylight is tens of thousands, an office about 500, a street at night a few, moonlight under one. Low light.
Minimum illumination: The lowest light at which a datasheet claims a usable image, under test conditions it often does not spell out.
IR range: How far the camera's built-in infrared lamps reach. Beyond it the night image is dark, however good the sensor.
Wide dynamic range (WDR): Combining exposures, or a sensor built for contrast, so bright windows and dark rooms read in one frame. Quoted in dB. Wide dynamic range.
Thermal camera: A camera that images long-wave infrared, heat, instead of light. Detects at long range in darkness; does not identify people.

Dome, turret, bullet: The three common fixed housings: a dome under a clear cover, an eyeball turret, and a cylindrical bullet. Camera form factors.
PTZ: Pan, tilt, zoom: a motorised camera that can be steered. Covers a lot of area, but only one part of it at a time.
Multi-sensor camera: Several sensors in one housing, each with its own view, usually aimed to cover 180° or 360° from one mount.
NVR / DVR: Network or digital video recorder: the box that records the cameras. Sized by channels, storage and bitrate. Storage and retention.
IP and IK ratings: Ingress protection against dust and water (IP66, IP67) and impact resistance against vandalism (IK10).

Retention: How many days of recording the system keeps before overwriting. Storage equals bitrate × time × cameras, with a margin.
PoE: Power over Ethernet: the camera takes its power down the network cable. Classes set the power budget per port. PoE and cable runs.
The 100 m rule: The maximum length of a copper Ethernet run between switch and camera, including patch leads.
ONVIF: An industry standard that lets cameras and recorders from different makers work together. Profiles S, G and T cover streaming, recording and modern codecs.
H.264 / H.265: The video codecs nearly every camera uses. H.265 roughly halves the bitrate for the same picture.

Coverage: The union of floor that at least one camera sees, after walls and obstacles have cut every cone.
Blind spot: Floor no camera sees, including the strip under each wall-mounted camera. Mounting height and the blind spot.
Occlusion: Walls, columns and shelving blocking a camera's view. CameraCAD ray-casts every cone against what is drawn.
Underlay: A scanned plan, PDF or photo placed under the drawing to trace over. Reading a floor plan.
Scale calibration: Setting an underlay's true scale by marking two points a known distance apart.
Camera schedule: The table of every camera on a drawing, with model, settings, mount and cable. The camera schedule.
Title block: The box on a printed drawing naming project, author, scale, date and revision.

EN 62676-4: The part of the international video surveillance standard dealing with application guidelines, including the DORI densities.
NDAA Section 889: The US law barring federal purchase of video surveillance equipment from named manufacturers, and federal contracts with anyone who uses it. NDAA compliance.
OEM / relabelled camera: A camera made by one manufacturer and sold under another's brand. Compliance follows the maker, not the label.