🎨 Poly Art Generator

Transform Images into Low-Poly Artwork with Computational Geometry

A Python-based tool that converts any input image into stunning low-poly art using computational geometry and advanced image processing techniques.

---

🚀 Overview

Poly Art Generator is a geometry-driven image processor that transforms standard images into stylized polygon meshes. Perfect for digital art, posters, NFTs, and creative coding projects. Unlike simple filters, this tool uses real computational geometry to create authentic low-poly artwork.

✨ Key Features

• 🎯 Geometry-Driven Processing - Uses Delaunay triangulation, not basic filters
• 🎨 Customizable Polygon Density - Control the level of detail in your artwork
• ⚡ Lightweight & Offline - No internet connection required
• 🔧 Extensible Architecture - Easy to modify and enhance
• 📦 Production Ready - Clean code with modular design

---

🧠 How It Works

Image Preprocessing Resize & optimize input

Point Sampling Strategic point distribution

Triangulation Delaunay mesh generation

Color Mapping Average color per triangle

Rendering Generate final artwork

---

💻 Tech Stack

Python 3.x

OpenCV

NumPy

SciPy

Pillow

---

⚙️ Installation

# Clone the repository
git clone https://github.com/your-username/poly-art-generator.git

# Navigate to project directory
cd poly-art-generator

# Install dependencies
pip install -r requirements.txt

📋 Requirements

opencv-python>=4.5.0
numpy>=1.19.0
scipy>=1.5.0
Pillow>=8.0.0
matplotlib>=3.3.0

---

🎯 Quick Start

from poly_art import PolyArtGenerator

# Initialize the generator
generator = PolyArtGenerator()

# Convert your image
generator.generate(
    input_path="input/photo.jpg",
    output_path="output/poly_art.jpg",
    num_points=500  # Adjust for polygon density
)

📂 Project Structure

poly-art-generator/
│
├── poly_art.py           # Main generator script
├── requirements.txt      # Project dependencies
├── README.md            # Documentation
│
├── input/               # Place your input images here
│   └── sample.jpg
│
├── output/              # Generated artwork saves here
│   └── result.jpg
│
└── examples/            # Sample outputs & demos
    ├── landscape.jpg
    └── portrait.jpg

---

🎨 Usage Examples

Basic Usage

# Simple conversion with default settings
python poly_art.py --input input/photo.jpg --output output/result.jpg

Advanced Configuration

# Fine-tune polygon density
python poly_art.py \
    --input input/photo.jpg \
    --output output/result.jpg \
    --points 1000 \
    --quality high

Batch Processing

# Process multiple images
python poly_art.py --batch --input-dir input/ --output-dir output/

---

🎬 Results Showcase

Original Image	Low-Poly Art (500 points)	Low-Poly Art (1000 points)
Your before image here	Your after image here (low density)	Your after image here (high density)

---

🛠️ Configuration Options

Parameter	Type	Default	Description
num_points	int	500	Number of sampling points (controls polygon density)
quality	str	'medium'	Output quality: 'low', 'medium', 'high'
max_size	int	1920	Maximum dimension for processing
edge_detection	bool	True	Enable edge-aware point sampling

---

🔬 Algorithm Details

Point Sampling Strategy

The generator uses intelligent point sampling that focuses on:

• Image edges and high-contrast regions
• Uniform distribution across flat areas
• Corner preservation for structural integrity

Delaunay Triangulation

Employs SciPy's spatial Delaunay implementation for:

• Optimal triangle mesh generation
• No overlapping polygons
• Mathematically sound geometry

Color Averaging

Each triangle color is computed using:

• Mean RGB values within triangle bounds
• Optional weighted averaging based on pixel importance
• Anti-aliasing for smooth color transitions

---

🚀 Performance

• Speed: Processes 1920×1080 images in ~2-5 seconds
• Memory: Peak usage ~200MB for high-resolution images
• Scalability: Handles images up to 4K resolution

---

🤝 Contributing

Contributions are welcome! Here's how you can help:

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

---

📝 Roadmap

• Add GUI interface
• Implement color palette customization
• Support for video processing
• Export to SVG format
• Real-time preview mode
• GPU acceleration with CUDA

---

🐛 Known Issues

• Very high polygon counts (5000+) may cause performance issues
• Transparent PNGs are converted to RGB format
• Some EXIF metadata is not preserved

---

🙏 Acknowledgments

• Delaunay triangulation algorithm by SciPy
• OpenCV community for image processing tools
• Inspiration from low-poly art movement

---

📧 Contact

---

⚡ "Art is not what you see, but what you make others see." - Edgar Degas

Made by Aryan Ranjan | Star ⭐ this repo if you found it helpful!