Privacy Masking
Privacy masking allows you to define areas in a camera's view that should be excluded from AI detection. By drawing shapes over specific zones, you prevent the system from raising alarms for activity in those areas, reducing false alarms from irrelevant activity and ensuring you do not inadvertently monitor areas that should remain private.
Why Use Privacy Masks?
Privacy masks are useful in several situations:
- Public roads and sidewalks -- If your camera captures a portion of a public road, mask it to avoid generating alerts from passing pedestrians and vehicles.
- Neighbor's property -- Mask any areas that overlap with neighboring properties to respect their privacy.
- Windows and glass surfaces -- Reflections and movement visible through windows can trigger false detections. Masking windows helps prevent this.
- Trees and bushes -- Vegetation moving in the wind is a common source of false motion detections. Mask heavily vegetated areas if they cause problems.
- High-traffic zones -- Areas where routine activity is expected (for example, a busy walkway during business hours) can be masked to focus detection on more important zones.
The Shape-Based Mask Editor
Privacy areas are drawn as shapes with adjustable corners over the camera's snapshot. Each shape is a separate area (up to 8 per camera), shown as a dimmed veil labelled Not detected. The editor opens full-screen in landscape.
- Move a corner -- Drag any solid corner handle.
- Add a corner -- Tap the small hollow handle in the middle of any edge; a new corner appears there. A shape can have up to 64 corners.
- Remove a corner -- Tap a corner to select it, then tap Remove corner. A shape always keeps at least 3 corners.
- Move a whole shape -- Drag anywhere inside the shape.
- Snap to the frame edge -- Corners dragged close to the edge of the picture snap onto it, so it is easy to mask right up to the border.
Working with multiple areas
A row of chips (Area 1, Area 2, ...) selects between shapes -- or just tap a shape directly.
- Add area -- Adds a new rectangular shape you can then reshape.
- Remove area -- Deletes the selected shape.
- Clear -- Removes every shape from the camera.
Tap Done to save, or Cancel to discard changes. The editor will not save if the shapes cover almost the entire picture (over 90%) -- shrink an area or use Clear instead of masking everything.
How to Create a Privacy Mask
- Open the camera and, in the Privacy zones card, tap Set privacy region (also available in the camera setup wizard).
- The editor opens over the camera's current snapshot.
- Tap Add area and drag the corners of the new shape over the zone you want detection to ignore -- like a public pavement. Tap edge midpoints to add corners for finer outlines.
- Add more areas if needed, then tap Done.
Editing an Existing Privacy Mask
Open the editor the same way -- the camera's current shapes load ready for adjustment.
If the camera's mask was drawn with the older grid-based editor, it is converted to shapes automatically when the editor opens. In rare cases where the conversion cannot reproduce the old mask exactly, the editor shows "Converted from the old grid -- check the shapes before saving." -- review the shapes before tapping Done. Saving stores the mask in the new shape format from then on.
How an Object Is Judged Against a Privacy Area
A privacy area is judged at the bottom fifth of a detected object -- a person's feet, a vehicle's wheels. At least half of that strip must lie outside every privacy area, otherwise the object is not a detection at all.
- Someone standing inside a privacy area -- a pedestrian on a masked pavement, a car on a masked road -- raises nothing.
- Someone on your property whose head or shoulders merely overlap a privacy area behind them is still detected, and is shown whole: the blur is cut away around them so you can see who it is.
The blur on event photos and snapshots follows the shapes as you drew them.
On a camera connected to a CleverCam hub, the zone that is ignored currently reaches slightly beyond the shape you drew -- up to about 6% of the picture past its edge -- so a person standing right beside a privacy area can be missed. Keep privacy areas clear of gates, doors and paths by at least that much. A hub update that judges objects against the shape exactly as drawn is on its way; CleverMail and Dahua Direct cameras already do.
Removing a Privacy Mask
- Open the camera's privacy mask editor.
- Tap Clear, then Done.
Best Practices
- Be precise -- Mask only the areas you need to exclude. Over-masking reduces the effective detection area of your camera.
- Review regularly -- If your environment changes (new construction, removed trees, adjusted camera angle), review and update your masks.
- Test after masking -- After saving a privacy mask, monitor the events from that camera for a few days to confirm the mask is working as expected and genuine detections are still being captured.
- Do not mask entry points -- Avoid masking areas that cover doors, gates, or other entry points to your property, as these are the most critical detection zones.
Privacy masks affect alarm decisions, and masked regions are blurred on event photos and camera snapshots so operators and the control room never see inside them. The live video feed itself is not altered -- if you need the raw stream obscured (for GDPR compliance or similar requirements), consult your installer about privacy zone features available on the camera itself.
Masked areas will not generate any detections, regardless of how sensitive the camera is set. Make sure you do not accidentally mask areas that need to be monitored for security.