Spaces:
Sleeping
title: Smart Crowd Detector
emoji: 🏃
colorFrom: blue
colorTo: green
sdk: docker
pinned: false
app_port: 7860
Smart Crowd Detector
An AI-powered real-time crowd detection and monitoring system using YOLOv8 with an adaptive heatmap visualization. Built for safety, efficiency, and smart crowd management.
🎯 Features
- Real-time People Detection - GPU-accelerated YOLOv8 model with TensorRT, OpenCV, and PyTorch.
- Dynamic Detection Modes - Switch dynamically depending on your environment:
- 🟢 Stadium Mode: High density optimization (500 max detections, low confidence/IOU thresholds) perfectly tailored for massive crowds and distant individuals.
- 👥 Normal Mode: Balanced accuracy and framing for standard environments (offices, retail).
- ⚡ Fast Mode: Maximum framerate via skipped frames and lower resolutions for basic needs on lower-end hardware.
- Adaptive Heatmaps - Smart kernel sizing based on object distance to represent crowd density accurate to visual depth.
- Dual Source Input - Works seamlessly with an active webcam or video files with continuous looping playback support.
- Alert System - Highly configurable warning/critical crowd density thresholds.
- High Performance - Thread-safe state caching, frame deferring, GPU pipeline offloading, and optimized resolutions delivering 30-35 FPS on fast mode or extreme accuracy insights for dense stadium applications.
- Modern Dashboard - Clean F1-themed interface built on vanilla JS and standard web sockets.
📋 Requirements
- GPU: CUDA-capable NVIDIA GPU (RTX 3050+) is heavily recommended for Stadium mode.
- RAM: 8GB minimum, 16GB recommended.
- Software: Python 3.10+, CUDA 12.0+
🚀 Installation & Local Setup
# 1. Clone the Repository
git clone https://github.com/nowayitsme-eng/Smart_Crowd_Detector.git
cd Smart_Crowd_Detector
# 2. Create and Activate Virtual Environment
python -m venv venv
# On Windows:
venv\Scripts\activate
# On Linux/macOS:
source venv/bin/activate
# 3. Install Dependencies
pip install -r requirements.txt
# 4. Run the application
python app.py
Access the dashboard at http://localhost:5000
🐳 Docker & Cloud Deployment
⚠️ Note on Serverless (e.g. Vercel): Standard serverless does not support this application. Real-time video processing requires persistent connections, background sockets, and large ML sizes which bypass standard serverless constraints. Use Docker or standard scalable containers.
Docker (Recommended for AWS, GCP, Azure, DigitalOcean)
Create a Dockerfile with the following:
FROM python:3.11-slim
RUN apt-get update && apt-get install -y libgl1-mesa-glx libglib2.0-0 libsm6 libxext6 libxrender-dev && rm -rf /var/lib/apt/lists/*
WORKDIR /app
COPY requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt
COPY . .
EXPOSE 5000
CMD ["python", "app.py"]
Build and run:
docker build -t app-monitor .
docker run -p 5000:5000 --device=/dev/video0 app-monitor
Render.com / Railway.app
- Connect your GitHub Repository natively.
- Railway: Instantly supported utilizing persistent Docker builds.
- Render: Use a standard Web Service provider. Build command:
pip install -r requirements.txt. Start command:python app.py.
Heroku
Use Heroku buildpacks if not building via container setup:
heroku create smart-crowd-monitor
heroku buildpacks:add --index 1 heroku-community/apt
heroku buildpacks:add --index 2 heroku/python
git push heroku main
⚙️ Configuration
Tweak main properties centrally in config.yaml:
video:
source: 0 # Camera index or path/to/video.mp4
fps: 30
resolution:
width: 640
height: 480
model:
confidence_threshold: 0.35
device: "cuda" # switch to "cpu" if no GPU available
crowd:
density_threshold: 20
warning_threshold: 35
📄 License & Credits
MIT License
Author: Ali Abdullah - GitHub: nowayitsme-eng
Acknowledgments: YOLOv8 by Ultralytics, OpenCV, Flask Framework