People counting tracks how many people enter and exit each area, building up historical data on footfall patterns, peak hours, and occupancy levels. Heatmaps show where visitors spend time. Occupancy alerts fire when configured thresholds are approached or exceeded.
Counting quality depends on a clear overhead view at a single doorway. Open floors, crossing paths, and crowding make counts less reliable. Common applications include retail staffing, occupancy alerts, and entrance trends.
Stop guessing how many people visit your locations. AI-powered counting gives you accurate foot traffic data, occupancy monitoring, and trend analysis — all from your existing cameras on the edge box.
Deploying in a shopping center? See the mall people counting application.
How organizations turn foot traffic into actionable business insights
Combine foot traffic data with POS transactions to calculate true conversion rates. Know what percentage of visitors actually buy something.
Monitor real-time occupancy for fire code compliance, energy management, and space utilization. Know which floors and zones are busiest.
Track attendance at events, conferences, and venues. Monitor crowd density in real-time and manage capacity limits proactively.
From raw headcounts to predictive staffing and occupancy compliance
Track people entering and exiting through doorways and entrances. Know exactly how many people are inside at any given moment.
See where people spend time with color-coded heatmaps. Identify high-traffic zones, dead spots, and popular areas at a glance.
Identify your busiest hours, days, and seasons. Plan staffing, promotions, and operations around actual traffic patterns.
Set maximum capacity thresholds. Get automatic alerts when occupancy approaches or exceeds your limits.
Compare foot traffic week-over-week, month-over-month. Spot trends, measure the impact of events, and forecast future traffic.
Automated daily and weekly reports delivered to your inbox. Key metrics, trends, and anomalies summarized for quick review.
Counting accuracy is typically above 95% at entrances with clear overhead or angled views — crowded open spaces are significantly harder for any AI to track. For the best results, we recommend ceiling-mounted cameras at entry/exit points rather than trying to count people in wide open areas. Our team can help you identify the best camera positions.
At entrances with clear overhead or angled views, counting accuracy is typically above 95%. The AI uses DeepSORT tracking with Kalman filtering to avoid double-counting — it tracks individuals as they cross a line, not just detections per frame. Accuracy drops in crowded open spaces where people overlap and occlude each other. Narrow doorways and corridors give the best results.
The ideal setup is a camera mounted directly overhead (ceiling mount) at an entrance or exit point. This gives a clear view of each person crossing without occlusion issues. If overhead isn't possible, an angled view from 10-12ft height looking down at a doorway also works well. Side-angle views at ground level are the hardest — people occlude each other constantly.
Yes. You can set maximum occupancy thresholds per zone or per building. When the count approaches your limit, you'll get a warning alert. When it hits the limit, you'll get a critical alert. This is especially useful for fire code compliance, capacity limits, or event management where you need to control how many people are inside.
We retain counting data for the duration of your storage plan — typically 30, 60, or 90 days of detailed per-minute counts. Aggregated hourly and daily summaries are kept for up to a year. You can export data to CSV at any time for long-term archival or integration with your business intelligence tools.