Person Detection and Counting System for Public Transport

Objective:

This project aims to develop a real-time system that detects and counts people boarding a bus using advanced object detection techniques. The system leverages YOLO (You Only Look Once) for person identification and tracks individuals to ensure accurate counting, improving the management of passenger capacity in public transportation.

Description:

The project integrates a computer vision model trained on the COCO dataset with real-time video analytics to monitor and count passengers entering a bus. The key functionality includes:

1. Object Detection:
   • Uses the YOLOv8 model to identify and classify objects in video footage.
   • Focuses specifically on detecting people among other potential objects like bags, bicycles, or vehicles.
2. Passenger Counting:
   • Counts individuals when they cross a designated horizontal line (representing a boarding threshold) on the screen.
   • Ensures accurate counts by tracking unique individuals using bounding box properties.
3. Real-Time Processing:
   • Processes video frames efficiently to handle real-time requirements.
   • Displays the detected people and the total count on the screen while printing updates in the terminal.
4. Application Context:
   • The system is tailored for use in buses to count passengers boarding. By tracking people crossing a specific area, it prevents double counting and provides accurate statistics.

Use Case:

The project can be deployed in surveillance systems within buses or at bus stops to monitor passenger flow, optimize bus schedules, and enhance public transport efficiency. For instance, it can provide insights into the busiest routes or times by analyzing historical data.

Key Features:

• Unique Tracking: Uses bounding box data to uniquely identify individuals and prevent duplicate counts.
• Customizability: Can detect and count other objects by extending the functionality to additional classes in the COCO dataset.

Conclusion:

The person detection and counting system effectively utilizes YOLO for accurate real-time passenger monitoring. It ensures reliable counting by tracking individuals crossing a defined threshold and displays the count visually on-screen and in the terminal. This project demonstrates significant potential for enhancing public transport management and safety compliance.