People Counting
Description
This metric enables accurate counting of people crossing a defined point, distinguishing between entries and exits. It is essential for occupancy analysis, flow monitoring, and retail behavior insights.
| Element | Detail |
|---|---|
| Description | Measures how many people cross a virtual line (door, aisle, checkout), distinguishing Entries and Exits. |
| Algorithm | 1. Detection: AI anonymizes and detects people. 2. Tracking: generates unique trajectories. 3. Intersection: when the trajectory crosses the line, it is counted as +1 Entry or Exit. |
| Key Parameters | • Line thickness (px) • % crossing validation • Enabled direction (Entry, Exit, both) |
| Accuracy | Accuracy > 99 % |
| Typical Use Cases | • Visitors • Store occupancy • Mall capture • Multi-door flow balance |
Visual Explanation
| Visual Explanation |  |  |  |
|---|---|---|---|
| Description | 1. Detection: AI anonymizes and detects people. | 2. Tracking and intersection counting: unique trajectories are generated, and when the trajectory crosses the line, it is counted as +1 Entry or Exit. | Key Parameters: the trajectory must exceed a % of the line thickness to be counted. |