Mall People Counting
Entrance footfall, floor occupancy, and heatmaps on the AI Edge Analytic Box. Designed for doorway accuracy and fire-code occupancy limits.
Multi-entrance counting
Separate lines per gate, anchor store, and parking structure.
Zone heatmaps
See where visitors linger — food court vs. luxury wing.
Peak-hour analytics
Compare Saturdays, holidays, and campaign weekends.
Occupancy caps
Local alerts when floors approach capacity limits.
FAQ
How accurate is counting at mall entrances?
With overhead or high-angle doorway views, entry/exit counting typically exceeds 95%. The edge box uses ByteTrack-style tracking to avoid double-counts when crowds bunch at doors. Open atrium views are harder — we recommend dedicated counting lines at each entrance.
Can we see occupancy by floor or zone?
Yes. Assign counting lines per entrance and zone polygons per floor. The dashboard shows live occupancy, peak hours, dwell heatmaps, and historical trends. Occupancy cap alerts fire locally before cloud sync.
Does counting run without internet?
All inference and hourly rollups run on the AI Edge Analytic Box. Internet is only needed if you want centralized dashboards across multiple malls — counts buffer locally until sync.
How does this relate to the People Counting AI page?
The mall application is our retail-footfall deployment guide — same Vision Engine models, packaged for shopping centers with multi-entrance layouts. Technical deep-dive: /ai/people-counting.