How Many Security Cameras Does a Building Need?

Start from purposes, not from square footage#

The count comes from what someone will need to see, from where, and at what level of detail. Square footage does not produce that answer, and neither does a rule of thumb per floor or per door.

Write a purpose for every camera before anything is placed on a plan. A purpose is one sentence with a verb, a subject and a location: identify the face of anyone entering through the main lobby door, detect a person crossing the north fence line, read the plate of every vehicle entering the parking structure, see who handles packages at the mail room counter. If a proposed camera cannot be given a sentence like that, it does not belong on the plan, and if a stated purpose has no camera assigned to it, one is missing.

This is not a documentation exercise. The purpose determines resolution, lens, mounting height, aiming angle, whether infrared or wide dynamic range is needed, and how much storage the camera consumes. It also determines whether the system is useful after an incident. Most disappointing camera systems have plenty of cameras and no stated purposes, so every view is a compromise between two things nobody chose.

The purpose list also survives the project. Handed over as a camera schedule, it tells the next technician why a camera is aimed the way it is, and it tells the owner what the system was designed to do when someone asks why an incident was not captured.

Pixels on target, lens choice and mounting height#

Megapixels describe the sensor. Pixel density describes what actually reaches the subject, and it is the number that decides whether footage is usable. Pixel density is horizontal resolution divided by the width of the scene at the target distance, expressed in pixels per metre or pixels per foot of target.

The DORI criteria in IEC 62676-4 give four thresholds, and they are the values manufacturers publish in their own coverage tools.

Written purpose	DORI level	Pixel density on target	Typical lens and placement
Identify an unknown face at a controlled entry	Identify	About 250 px/m, roughly 76 px/ft	Narrow view about 8 to 12 ft wide, camera at 7 to 8 ft aimed at head height
Read a plate at a gate or lot entry	Identify, plate specific	Above 250 px/m on the plate itself	Purpose built plate camera with shutter and infrared tuned for retroreflective plates
Recognize a known employee in a corridor	Recognize	About 125 px/m, roughly 38 px/ft	Fixed view down the length of the corridor, not across it
Observe activity at a counter or dock	Observe	About 62.5 px/m, roughly 19 px/ft	Moderate view, 15 to 25 ft wide, mounted to avoid backlight
Detect a person along a fence or in a lot	Detect	About 25 px/m, roughly 8 px/ft	Wide view at longer range, paired with lighting or infrared

Two consequences follow. First, a wide lens spreads a camera’s pixels thin. An 8 MP camera puts about 3,840 pixels across its image, and a 110 degree lens spreads them over roughly 114 feet of scene width at 40 feet of range, which is about 34 pixels per foot. That is enough to observe activity, short of the 38 px/ft that recognizes a known person, and well short of the 76 px/ft that identifies a stranger. Put a 30 degree lens on the same sensor and the view narrows to about 21 feet at that range, raising density to roughly 180 pixels per foot, which identifies. Second, mounting height changes what the camera sees more than resolution does. A dome at 14 feet in a corridor shows the tops of heads. The same camera at 8 feet, aimed along the corridor at head height, shows faces. Ceiling height is a design constraint, not a detail.

Aim matters as much as height. A camera facing a glass entry door with daylight behind it produces silhouettes unless wide dynamic range is specified and the camera is placed to look at arriving faces rather than into the sun. Our camera coverage calculator works the resolution, field of view and distance tradeoff for a quick check before a site walk.

Positions almost every commercial building needs#

The following positions appear on nearly every commercial camera plan, and each of them exists because something specific happens there.

• Every exterior door, including stair discharge and service doors. Doors are where identity is established and where forced entry happens. Aim for identify at the primary entrance, recognize at the rest.
• Lobby and reception, one overview covering the space and the path to the elevators or the corridor.
• Elevator lobbies on every floor, and inside each cab. Cab cameras are the only view of what happens between floors.
• Corridor decision points, meaning junctions where a person can go more than one way. Cameras mid corridor add little; cameras at junctions let a reviewer follow a path.
• Parking entries and drive aisles. Entries get identify or plate cameras. Aisles get detect coverage.
• Loading dock and receiving. Both the door and the area where goods sit before they are moved.
• Mail and package room. Package theft is the most commonly reviewed footage in multifamily and office buildings.
• Cash handling and high value storage, wherever those exist.
• Telecom and mechanical rooms. A camera on the door of the MDF documents who touched the network and the recorder.
• Building perimeter and blind corners, where the building’s shape creates approaches nobody can see from inside.

Cameras work alongside doors, not instead of them. A camera records that a door opened; an access control system records who opened it and can stop them. Tying the two together so a forced door alarm brings up the clip is a configuration step, not an assumption.

How building type changes the count#

Building type	Positions that drive the count	Additions beyond the standard set
Multifamily	Every building entry, mail and package room, amenity rooms, garage entries, elevator lobbies and cabs, trash enclosure	Bicycle storage, play areas, gated vehicle entry; no view into any unit
Senior housing	Same as multifamily, plus every controlled egress door	Doors with wander risk, dining room entry, medication storage; corridors sized for recognize
Office tenant improvement	Suite entry, reception, server room, tenant corridor junctions	Base building may already be covered by the landlord; confirm before duplicating
K-12 and higher education	Every exterior door, front office and visitor vestibule, cafeteria, gym entry, bus loop, parking, corridor junctions	Fence lines, portables, athletic facilities; the district or campus standard usually governs
Clinic	Waiting room, reception, controlled clinical entry, medication storage, staff entry, parking	Never in exam rooms; check facility policy before placing cameras in treatment corridors
Warehouse	Every dock door, shipping and receiving desks, yard entries, high value cage, employee entrance	Aisle overviews, trailer yard, charging and fueling areas

The pattern across building types is that count follows doors, transitions and assets. A building with one entrance and a large open floor needs fewer cameras than a smaller building with six doors and a dozen controlled spaces. Occupancy also matters: in multifamily and affordable housing, common area coverage is expected and unit level coverage is prohibited, which caps the count in a way an office building does not.

Multisensor and fisheye as count reducers#

Two form factors reduce device count when the geometry cooperates.

A multisensor camera puts three or four independently aimed sensors in one housing on one cable and one mount. At a building corner it covers two facades and the approach between them. At a corridor intersection it covers three or four legs. Where it backfires: when the ideal mounting point is not where the views converge, when the sensors cannot be aimed steeply enough for the ceiling height, when the added weight needs a mount the existing structure cannot take, and when the platform licenses per sensor rather than per device, which erases the savings on a cloud subscription. Ask how the platform counts a multisensor before assuming a four sensor unit costs one license.

A fisheye covers 360 degrees from a ceiling and dewarps in software. It works in small lobbies, break rooms, package rooms and small retail floors, where nothing important happens more than 15 or 20 feet from the mount. Detail falls off toward the edge of the image, so a fisheye covering a 40 foot room is a presence detector at the walls, not an identification camera.

Both reduce cable drops, which is where a meaningful part of the installed cost lives. Neither reduces storage or bandwidth, because the pixels still have to be recorded. Our guide on where recorded video should live covers what that means for the recorder.

Worked example: a small office building#

Assumptions: two stories, one main lobby entrance, one rear employee entrance, one service door at a small dock, two stair discharge doors, one elevator serving both floors, one surface parking lot with two vehicle entries, an MDF on the ground floor and an IDF upstairs, and a mail and package room off the lobby.

Written purpose	Position	Cameras	Level
Identify anyone entering the building	Main entry, inside the vestibule at head height	1	Identify
See who is in the lobby and where they went	Lobby overview	1	Observe
Identify staff using the rear door	Rear employee entrance	1	Identify
Identify deliveries and after hours access	Service door and dock	1	Identify
Confirm the stair exits are not propped or used for entry	Exterior of each stair discharge door	2	Recognize
See who boards and leaves the elevator	Elevator lobby, each floor	2	Recognize
See what happens between floors	Elevator cab	1	Recognize
Follow movement past the corridor junctions	Second floor corridor decision points	2	Recognize
Read plates and see drivers entering the lot	Each vehicle entry	2	Identify
Cover the drive aisles and stalls	Lot overview	2	Detect
See who enters the telecom rooms	MDF and IDF doors	2	Recognize
See package handling	Mail and package room	1	Observe
Cover the two approaches nobody can see from inside	Exterior building corners	2	Detect

That totals 20 cameras, each with a sentence attached. Substituting multisensor units at the two exterior corners keeps the device count at 20 but delivers six views instead of two, at the cost of two heavier mounts and, on some platforms, six licenses. Dropping the second floor corridor cameras and the IDF door takes it to 17 and gives up the ability to follow a path upstairs. Adding a second lobby view for wide dynamic range against the glass entry takes it to 21. Those are real decisions, and the purpose list is what makes them decisions rather than guesses.

Where cameras must not go#

Cameras belong where people do not have a reasonable expectation of privacy: entries, corridors, lobbies, parking, docks, mail rooms, elevators, mechanical and telecom rooms, and building perimeters. They do not belong in restrooms, locker rooms, changing areas, or aimed into any dwelling unit.

In multifamily and senior housing this extends past the obvious. A corridor camera positioned so a unit door swing shows the interior, or an exterior camera whose field of view includes a unit window, creates exposure the owner does not want and residents will object to. Document the field of view of every camera in the closeout package so the owner can answer that question with a drawing rather than a conversation.

Microphones are a separate question with a separate answer. California treats the recording of private conversations differently from video and consent rules apply, so microphones are commonly left disabled on commercial camera systems unless counsel has reviewed the intended use. Signage stating that the property is under video surveillance is a sound practice and some owner policies and insurers require it. Our security camera systems page covers how those settings are documented at closeout.

Next step#

Send a floor plan and a list of what you need to see, or just the list, and we will return a camera schedule with a written purpose, a level and a lens for each position. Start at request a bid. General contractors bidding Division 28 can use for general contractors for our prequalification packet.