People Counter for Events

Advanced AutonomyIntermediate Robotics plus perception pipelines, control loops, coordinate frames, and real-time behavior.

Before you start

  • Complete Camera Module
  • Understand object tracking, image coordinates, and basic data logging
  • Use an instructor-approved person detector

Build a stationary camera project that estimates how many people cross a doorway or hallway boundary during a session.

Project Outcome

Display a live in, out, and total count with a timestamped session summary.

Camera Input

  • Person detections and stable track identifiers
  • A virtual crossing line in image or depth space
  • Recent positions for each active track

Build Stages

  1. Draw and verify the crossing line.
  2. Detect people without counting them.
  3. Track each detection across frames.
  4. Count only when a track crosses from one side to the other.
  5. Add a cooldown or completed-track record to prevent duplicates.
  6. Save aggregate counts without storing identifying imagery.

First Prototype

Count crossings in one direction and display the current total.

What To Measure

  • Missed crossings
  • Duplicate counts
  • Behavior when people overlap or reverse direction
  • Count accuracy over a known test group

Extensions

  • Add separate entry and exit totals.
  • Show hourly counts on a local dashboard.
  • Compare image-line and depth-line crossing logic.
  • Add a reset control for a new session.

Safety And Privacy

Keep the robot stationary and position equipment so it does not block the walkway. Store aggregate counts by default; do not add face recognition or identifying records.