How Do Surveillance Cameras Work: A Complete Guide to Camera Surveillance Systems?

A surveillance camera captures light through a lens onto an image sensor, converts that light into a digital signal, compresses the video, and sends it to a recorder, monitor, or app for storage and viewing.

For the full breakdown, see our best Camera Surveillance System guide.

Every security camera you’ve ever walked past runs the same four-stage pipeline, whether it’s a $30 Wi-Fi door cam or a rack of commercial units wired into a building’s network closet. The parts change. The order never does. Knowing that order is what lets you diagnose a fuzzy image, understand why one camera records to a card and another needs a server, and tell the difference between a real hardware limitation and a settings problem. Here’s the full chain, plus what the standards actually require once a system is installed.

The Four Stages Inside Every Camera

Capture, convert, process, and move — that’s the whole job, and each stage has one component responsible for it.

Light enters through the lens, which focuses the scene onto an image sensor. That sensor is the heart of the camera, and it comes in two types: CMOS and CCD. Both turn incoming light into an electrical charge, but CMOS dominates modern gear because it’s cheaper to produce, draws less power, and packs the processing circuitry onto the same chip. CCD sensors still show up where image uniformity matters more than cost.

The sensor’s raw output is analog voltage. On a digital camera, that signal gets converted to digital data, then run through an image processor that handles exposure, white balance, noise reduction, and — critically — compression. Compression is what makes surveillance practical. Compression codecs cut that by orders of magnitude, which is why a modest hard drive can hold weeks of footage.

Once processed, the video has to go somewhere. That’s the transmission stage, and it’s where camera systems diverge most.

Where The Video Goes: Storage And Transmission Options

Processed video travels to one of several destinations, and the destination usually defines what kind of system you’re dealing with. A standard video surveillance system is built from three things per the published standards: image capture devices, the interconnections between them, and image handling devices.

Handling devices cover the range — a local monitor, a DVR, an NVR, a memory card, cloud storage, or a phone app. Older analog setups fed a DVR, which digitized and recorded the feeds. Modern IP cameras do the digitizing themselves and stream over Ethernet or Wi-Fi to an NVR or straight to the cloud. Power over Ethernet, or PoE, carries both data and power over one cable, which is why it’s the common choice for cameras that need to run continuously without a separate power drop at each mount point.

Remote viewing works because the camera or recorder exposes its feed over the network. A web browser interface handles desktop access; a mobile app handles the phone. The footage quality you see remotely depends on the upload bandwidth at the camera’s location, not the download speed on your end — a detail that explains most “why does my remote feed look terrible” complaints.

Retention comes down to a simple equation: storage capacity divided by bitrate. Cut the resolution, lower the frame rate, or trim the recording schedule, and the same drive holds more days.

If you’re weighing which of these pieces to buy together, our tested camera surveillance system roundup breaks down the kits that hold up in real use.

What The Standards Require After Installation

Getting a system running is the easy part. The published surveillance standards treat planning, testing, and documentation as mandatory, and that’s the part most home setups skip entirely.

The BS EN 62676 series, the European standard for video surveillance systems that’s widely referenced internationally, lays out the full lifecycle: risk assessment, operational requirements, a testing plan, site survey, system design, site plan, training, manuals, support documentation, installation details, and maintenance procedures. Every one of those is expected before and after deployment, not just at install.

The SWGDE recommendations for video security systems add documentation specifics that matter for anyone who might ever need footage as evidence. Keep the make and model of every component, the software version for digital systems, manuals, your retention schedule, passwords, and plainly written instructions for downloading or outputting recordings. That last item gets overlooked constantly, and it’s the one thing a person actually needs when there’s a deadline to hand over footage.

Two more caveats from the standards worth knowing. Version details matter because the software and hardware revision of a digital system affects how it operates and how evidence gets handled — so document them. And in one security-systems specification, cameras are required to be visible rather than hidden, with covert placement requiring formal approval. If you’re deploying cameras at a business, check that rule before mounting anything in a concealed spot.

The maintenance requirement is the one that separates a working system from a dead one. A camera that hasn’t been tested in two years is a camera you’re guessing about. Infrared-sensitive cameras in particular should have their locations documented and their maintenance records kept, since IR performance degrades quietly and nobody notices until they need night footage that doesn’t exist.

The Bottom Line

A surveillance camera is a light-to-data converter with a storage destination attached. The lens gathers, the sensor converts, the processor compresses, and the network delivers — and the standards governing professional installations demand that the whole chain be planned, tested, documented, and maintained rather than simply switched on. Get the four stages clear in your head and every decision about resolution, storage, and network setup gets easier.