CCTV (closed-circuit television) is a video surveillance system in which cameras send their footage over a private, closed network to recording and viewing equipment, instead of broadcasting it openly the way television does. A modern CCTV system has four parts: cameras that capture the scene, a network or cable run that carries the signal, a recorder (a DVR or NVR) that stores it, and a monitor or software client that lets a person watch live or search the archive.
Key takeaways
- CCTV is “closed” because the signal goes only to the equipment the owner connects, not to the public.
- Analog CCTV sends video over coaxial cable to a DVR. IP CCTV sends digital video over Ethernet to an NVR, and is the type most new installations use.
- Power over Ethernet (IEEE 802.3af supplies up to 15.4 W per port, 802.3at up to 30 W) lets one network cable power an IP camera and carry its video, with a standard run limit of 100 meters.
- Compression such as H.264 and H.265 decides how much storage a camera consumes, so retention time is a storage calculation, not a setting you can ignore.
- Plain CCTV records. It does not understand what it records, which is the gap video analytics is built to close.
What Does CCTV Stand For and How Does It Work?
CCTV stands for closed-circuit television. A camera converts light into a video signal, the signal travels to a recorder over a dedicated connection, and the recorder writes it to disk while also sending a live view to a screen. Because the circuit is closed, only devices on that network can see the video, which is the main difference between CCTV and a public broadcast or a consumer streaming camera that sends footage to an outside cloud by default.
The Four Parts of a CCTV System
- Cameras: fixed, dome, bullet, or pan-tilt-zoom (PTZ) units, chosen by the field of view and mounting position.
- Transmission: coaxial cable for analog systems, Ethernet (Cat5e or Cat6) for IP systems, or a wireless link for hard-to-wire locations.
- Recording: a DVR (digital video recorder) for analog cameras or an NVR (network video recorder) for IP cameras. The DVR vs NVR comparison covers how to choose between them.
- Viewing and storage: a monitor, a client application, and the disks or cloud storage that hold the archive.
Analog CCTV vs IP CCTV: What Is the Difference?
Analog CCTV sends a video signal over coaxial cable to a DVR that digitizes it. IP CCTV digitizes the video inside the camera and sends it over a standard network to an NVR, which allows higher resolution, power over the same cable, and software integration. Most new sites choose IP, while many older sites still run analog cameras that can be kept in service alongside new IP units.
| Feature | Analog CCTV | IP CCTV |
|---|---|---|
| Cable | Coaxial plus separate power | Cat5e or Cat6 Ethernet, power over the same cable with PoE |
| Recorder | DVR | NVR or server |
| Where video is digitized | In the recorder | In the camera |
| Maximum cable run | Depends on cable type and signal format | 100 m per Ethernet segment |
| Interoperability | Vendor and format specific | ONVIF Profile S and Profile T |
| Software analytics | Limited | Supported on the stream or on the camera |
Where Is CCTV Used?
CCTV is used wherever an owner needs a record of what happened on a site. Typical locations include retail floors and stockrooms, warehouses, parking lots, building entrances, construction sites, ports, and utility sites. The purpose differs by site: retail uses it for loss prevention and dispute resolution, industrial sites for safety and access records, and critical facilities for perimeter and incident review.
How Long Is CCTV Footage Kept?
Footage is kept for as long as the owner’s storage and policy allow, and there is no single universal retention period. Storage need depends on camera count, resolution, frame rate, compression, and how much of the day the camera records. A site that records continuously at higher resolution needs far more disk than one that records only on motion, and H.265 compression generally needs less storage than H.264 for the same image quality. Privacy rules such as GDPR in the EU also limit how long identifiable footage should be retained, which is why many operators set a fixed overwrite window instead of keeping everything.
What Affects Storage Needs
- Resolution and frame rate of each camera
- Compression codec (H.264 or H.265)
- Continuous recording versus motion or event recording
- Number of cameras and required retention window
What Types of CCTV Cameras Are There?
CCTV cameras are grouped by shape and by how they move. The right type depends on the area to cover, the mounting height, and whether the camera needs to be visible as a deterrent or discreet.
Common Camera Types
- Dome cameras: compact housings for ceilings and entrances, harder to tamper with because the viewing direction is hidden.
- Bullet cameras: long, visible units used outdoors on walls and poles, usually with infrared illumination for night use.
- PTZ cameras: pan, tilt and zoom units that cover a wide area and can follow a subject, at a higher cost per unit.
- Fisheye or panoramic cameras: one lens covering a full room or yard, with the image dewarped in software.
- Thermal cameras: detect heat instead of visible light, used for perimeters and low-light sites.
How Do You Choose a CCTV System?
Start from the job the footage has to do, then work backward to the equipment. A camera that must read a license plate needs a different lens, resolution and mounting angle than one that only confirms a person entered a room. Outdoor units are commonly rated IP66 or IP67 under IEC 60529, and choosing cameras that support ONVIF profiles keeps the site free to change recorder or software later. Before buying, confirm cable distances, power budget per PoE switch, storage for the required retention window, and who is allowed to view the footage. For footage that shows identifiable people, review the rules that apply in your region, for example GDPR Article 4 in the EU, before setting retention and access.
What Can CCTV Not Do?
A conventional CCTV system records video but does not interpret it. A person has to watch the screens or search the archive after an incident, and attention drops quickly when one operator watches many feeds. Basic motion detection adds alerts, but it triggers on shadows, weather, and animals, so many sites end up ignoring it. This is the practical limit that led to AI cameras and video analytics, which run software on the video to detect defined events and let staff search footage in plain language.
From CCTV to Video Analytics
Video analytics software adds understanding on top of the cameras a site already has. Instead of replacing CCTV, an analytics layer reads the existing streams, so a site keeps its cameras and recorders. Argu’s Vision Agents work this way: you describe in a sentence what you want watched, and the agent applies it to your own camera footage. The vision agents guide explains the approach in more detail.
Frequently Asked Questions
What does CCTV stand for?
CCTV stands for closed-circuit television. The video travels over a private circuit to the owner’s own recorder and monitors rather than being broadcast to the public.
Is CCTV the same as a security camera?
A security camera is one component. CCTV refers to the whole system: cameras, transmission, recorder, storage, and viewing software.
What is the difference between a DVR and an NVR?
A DVR digitizes and records the signal from analog cameras sent over coaxial cable. An NVR records digital video that IP cameras have already encoded, sent over a network.
Can CCTV work without internet?
Yes. A closed local network with cameras and a recorder works with no internet connection. Internet access is only needed for remote viewing or cloud storage.
Can I add AI to an existing CCTV system?
Often yes. Video analytics software can read the video streams of existing IP cameras that support a standard such as ONVIF, so many sites add detection and search without replacing their cameras.
Want to see AI analytics run on your own CCTV footage?
Last updated: October 2026
