Skip to content

Latest commit

 

History

History
118 lines (82 loc) · 3.14 KB

README.md

File metadata and controls

118 lines (82 loc) · 3.14 KB
Raw

Crowd Detection and Wait-time Predictions

This project demonstrates real-time video processing using the YOLOv8 object detection model to count people and simulate POS (Point of Sale) operations. The application features an interactive front-end to select video streams from multiple outlets and provides live updates on the detected people count, bill simulations, and estimated wait times.

Features

  • Real-time Video Analysis: Detect and count people in video streams using the YOLOv8 model.
  • POS Simulation: Simulates bill processing for a food outlet with dynamic wait time estimation.
  • Interactive UI: Users can select between different outlets and monitor the results in real-time.
  • Flask + SocketIO Integration: Seamless communication between the back-end and front-end for real-time updates.
  • Themed UI Design: Modern, responsive interface with a dark mode theme.

Getting Started

Follow these steps to set up and run the application on your local machine.

Prerequisites

  • Python 3.8 or later
  • Virtual environment (recommended)
  • A YOLOv8 model file (yolov8n.pt)

Installation

  1. Clone the repository:

    git clone https://github.com/your-username/live-yolov8-pos.git
cd live-yolov8-pos

  2. Install the required dependencies:

    pip install -r requirements.txt

  3. Set up pyngrok for public access (optional):

    • Install ngrok and set up your API key: 
      from pyngrok import ngrok
ngrok.set_auth_token("YOUR_NGROK_AUTH_TOKEN")

  4. Ensure the video files exist in the specified paths:

    • Update the video_paths dictionary in app.py with your video file paths.

Running the Application

  1. Start the Flask application:

    python app.py

  2. Access the application:

    • Locally: http://127.0.0.1:5000
    • Via ngrok (if configured): Check the terminal output for the public URL.

Usage

  1. Open the application in your web browser.
  2. Select an outlet from the homepage to start processing its video feed.
  3. Monitor:
    • People Count: Real-time count of detected individuals.
    • POS Simulation: Bill count and estimated wait time.
  4. Return to the homepage to select a different outlet.

Technical Details

YOLOv8 Model

  • This project uses the YOLOv8 model for object detection.
  • The model is initialized with: 
    model = YOLO('yolov8n.pt')

POS Simulation

  • The POS system simulates bill processing, where:
    • Base Wait Time: 2 minutes per person
    • Additional Time: 1.5 minutes per bill
    • Food Preparation Time: 7 minutes

Flask-SocketIO

  • Used for real-time communication between the server and the front-end to push frame updates and POS data.

Customization

  1. Video Streams:
    • Update the video_paths dictionary in app.py to point to your video files.
  2. Wait Time Calculation:
    • Modify constants wait_time_per_person, additional_time_per_bill, and food_prep_time.

Requirements

Create a requirements.txt file to manage dependencies:

Flask
Flask-SocketIO
numpy
opencv-python
pyngrok
ultralytics