When evaluating store performance and customer behaviour, one crucial metric is the number of Unique Visitors — the count of distinct individuals who enter a venue over a given period. At FootfallCam, this data is accurately captured by our flagship people counter, the FootfallCam 3D Pro2, using cutting-edge sensor fusion technology.
How We Measure Unique Visitors
Rather than simply tallying every entry, FootfallCam goes further by identifying repeat visits within the same day and filtering them out, ensuring that each individual is counted only once. This distinction provides a more meaningful insight into genuine customer reach, rather than inflating numbers with repeat traffic.
To achieve this, the FootfallCam 3D Pro2 uses Sensor Fusion, combining multiple technologies:
- Bluetooth Low Energy (BLE) sensing
- Wi-Fi probe request tracking
- 3D stereo vision with Re-Identification (Re-ID)
Each technology offers a unique perspective. BLE and Wi-Fi capture device signals (like mobile phones), while 3D stereo vision analyses visual patterns to distinguish between individuals.
Why Sensor Fusion Matters
Each sensing method has strengths and limitations. By fusing these inputs, FootfallCam cross-validates results, improving overall confidence in the Unique Visitor count.
Particularly, Wi-Fi and BLE sensing provide device MAC addresses. While the sample size (people carrying detectable devices) isn’t exhaustive, it is statistically representative enough to serve as a reliable benchmark. We use the MAC code matching proportion as a ground truth to calibrate and adjust the confidence levels of the 3D Re-ID system.
This clever approach ensures that the system remains accurate even as customer behaviours — such as device-carrying habits — change over time.
A Quick Word on Re-ID
Re-ID (Re-Identification) is the AI technology that allows the system to recognise whether a person seen earlier is the same individual returning later. It analyses 3D body shape, movement patterns, and visual features, rather than relying on facial recognition, making it privacy-safe.
At FootfallCam, Re-ID is enhanced by Reinforced Learning. This means the system is constantly learning from its environment: using MAC code matching feedback, it fine-tunes its algorithms automatically. As a result, the Re-ID AI model self-regulates, continually optimising itself for different lighting conditions, layouts, and customer flows — ensuring long-term accuracy without manual intervention.
The Result: A Smarter, More Reliable Metric
Thanks to sensor fusion, benchmark calibration, and AI reinforcement, FootfallCam provides retailers, transport hubs, and public venues with a truly dependable Unique Visitor metric. With this, businesses can better understand genuine customer reach, evaluate marketing effectiveness, and plan operational resources with confidence.
