• 

  starting

• 

  exit=0 entry=1, inside=1

• 

  exit=1, entry=1, inside=0

• 

  exit=1, entry=2, inside=1

Real-Time People Counter (Entered or Exits)

Inspiration

Managing crowd density and monitoring occupancy in real-time is crucial for safety, security, and space optimization. Inspired by the growing need for automated tracking solutions in places like malls, offices, and public events, we built an intelligent system to accurately count people entering and exiting a designated area.

What It Does

The Real-Time People Counter detects, tracks, and counts individuals as they move through entry and exit points. It:

• Uses YOLO (You Only Look Once) for real-time person detection.
  
• Tracks movement using OpenCV-based centroid tracking.
  
• Differentiates between people entering and exiting by defining virtual boundary lines.
  
• Provides accurate, real-time statistics for occupancy monitoring.
  

How We Built It

The system was developed using:

• Python as the primary programming language.
  
• OpenCV for image processing and motion tracking.
  
• YOLOv8 for fast and precise object detection.
  
• NumPy & Pandas for data handling and analysis.
  

Development Steps:

1. Capturing video feed from a live camera or pre-recorded footage.
   
2. Applying YOLO for object detection to recognize individuals.
   
3. Counting entries and exits by monitoring predefined regions.
   

Challenges We Ran Into

• Tracking objects across frames: Ensuring each person is correctly identified and followed between consecutive frames.
  
• Displaying real-time counts: Keeping an accurate and updated count of people entering and exiting.
  

Accomplishments That We're Proud Of

• Successfully implemented real-time tracking.
  
• Achieved high accuracy in entry and exit detection.
  
• Optimized performance to work with live camera feeds.
  
• Designed a scalable solution adaptable to multiple environments.
  

What We Learned

• How to integrate YOLO with OpenCV for efficient people tracking.
  
• Optimizing object detection models for real-world applications.
  

What's Next for Real-Time People Counter

• Web Dashboard Integration: Display real-time statistics on a user-friendly interface.
  
• Multi-Camera Support: Track people across different entry/exit points using multiple cameras.
  
• AI-Based Re-Identification: Improve tracking by identifying individuals across frames.
  
• Edge Device Deployment: Optimize the model for running on edge devices like Raspberry Pi for portable solutions.
  

Built With

• numpy
• opencv
• python
• yolov8