Write one measurable purpose for each scene—such as detecting presence, observing activity, recognizing a known person or supporting identification—then measure the target area and distance. Validate field of view, pixel density, mounting angle, day and night lighting, obstruction, compression, retention and export workflow on the real site. Higher resolution cannot compensate for the wrong lens, poor light, excessive angle or an over-wide scene.
Convert a vague goal into an operational requirement
| Purpose | What the image should support | Planning reference |
|---|---|---|
| Detect | Determine whether a person is present | About 25 pixels per metre |
| Observe | See number of people and characteristic details | About 63 pixels per metre |
| Recognize | Judge whether a person is the same as one seen before | About 125 pixels per metre |
| Identify | Support identifying a person in suitable conditions | About 250 pixels per metre |
Survey the real scene in day and night conditions
- Mark the exact target zone and purpose; an overview camera and an identification view are different jobs.
- Measure distance, target width, mounting options and likely approach direction rather than estimating from a floor plan alone.
- Check daylight, backlight, headlights, shadows, infrared reflection, weather, vegetation and seasonal obstruction.
- Calculate field of view and pixel density for the target zone, then validate on-site because compression, optics, angle and lighting also affect usable detail.
- Check power, PoE budget, cable path, surge and weather protection, network capacity, recorder channels, storage and export workflow.
- Confirm the precise interoperability features required by both client and device using the official ONVIF conformant-product records where applicable.
- Review privacy, notices, audio, neighbouring property, public areas, retention and access with the current applicable policy or legal guidance.
- Commission with recorded day/night samples and test playback, export, timestamps, account permissions and failure alerts.
Pixel density is a planning aid, not a guarantee
Axis explains that the pixel-density model simplifies a complex reality. Light direction, optics, compression, subject angle and display conditions can change whether an image meets the intended operational requirement. Test the real installation rather than treating a datasheet number as proof.
Security and privacy are part of scene design. Limit accounts and views to their purpose, use unique credentials, keep supported software current and avoid direct public-internet exposure of camera or recorder administration.
Sources and further reading
Technical statements were reviewed against these references. External pages may change after our review date.
- Pixel density based on IEC 62676-4:2014 — Axis Communications
- ONVIF Profiles — ONVIF
- Internet Exposure Reduction Guidance — Cybersecurity and Infrastructure Security Agency