Motion detection that fires on everything floods your alerts and gets ignored. The fix is tuning — combination of zones, sensitivity, and object filters.

1. Detection zones

Cameras typically let you draw zones where motion triggers (or doesn't):

  • Mask out areas with constant movement: trees, flags, busy roads in the distance, neon signs
  • Focus zones on areas you actually care about: doorways, parking spaces, gates
  • Use multiple zones with different sensitivities

2. Sensitivity

Too high: every shadow triggers. Too low: real intrusions missed.

  • Start at default (often 50%)
  • Walk through the scene yourself at various distances
  • Adjust until walking triggers but blowing leaves don't

3. Object detection (smarter cameras)

Modern cameras differentiate:

  • Human — most relevant for security
  • Vehicle — useful for driveways
  • Animal — useful to exclude (pets, wildlife)

Configure alerts for humans/vehicles only and ignore everything else. This dramatically reduces noise.

4. Line crossing

Instead of "motion in zone", set up:

  • A virtual line at a doorway or gate
  • Triggers only when something crosses the line
  • Direction can be specified (entering vs. leaving)

5. Loitering / intrusion

Advanced settings for high-security:

  • Loitering: alert when something stays in a zone for X seconds
  • Intrusion: alert if any movement enters a forbidden zone (like a server room outside hours)

6. Schedule

Different rules at different times:

  • 9-17: ignore most motion (office is busy)
  • 17-09: alert on humans only
  • Weekend: alert on any human

Most NVRs support schedule-aware motion detection. Configure to suit your operation.

7. Sensitivity changes with seasons

Outdoor cameras may need re-tuning:

  • Summer: more vegetation movement, more triggers
  • Winter: snow on a lens, ice on a connector
  • Spring: more wildlife, more false triggers

Quarterly review of motion settings is good practice.

8. Weather-based filters

Rain and snow create motion across the whole scene. Most cameras have:

  • "Rain/snow filter" or similar option that reduces sensitivity during precipitation
  • IR reflection issues — IR night vision can light up snowflakes near the lens, blinding the camera

9. Notification fatigue

If you've been getting notifications for months and ignoring them, the system has failed:

  • Either tune them down to actually-actionable events, or turn them off
  • Constant alerts you ignore = you'll miss the real ones

It's better to have NO notifications and check recordings periodically than have notifications you ignore.

10. Test after tuning

After making changes:

  • Stage a few scenarios you want to detect: walking, vehicle, loitering
  • Verify each triggers an alert
  • Stage scenarios you don't want: pets, wind in trees, passing cars at distance
  • Verify they don't trigger

When to ticket

Raise a ticket if:

  • The vendor's motion detection still produces too many false alarms after tuning
  • You need AI-based detection (vehicle, person, package) and your current cameras don't support it
  • You want a video analytics service (object counting, dwell time, etc.) configured

Include the camera models, what events you want to detect, and what's currently triggering false alarms.