🎨 EdgeMesh - Advanced 3D Mesh Generation from Images

Transform 2D images into stunning 3D meshes using AI-powered depth estimation and advanced edge detection ✨

🚀 Quick Start | 📸 Features | 🛠️ Installation | 📖 Documentation | 🎯 Examples

🌟 Overview

EdgeMesh is a powerful PyQt6-based desktop application that converts 2D images into 3D meshes using state-of-the-art depth estimation models and sophisticated edge detection algorithms. Whether you're a 3D artist, game developer, researcher, or hobbyist, EdgeMesh provides an intuitive interface for creating detailed 3D models from ordinary photographs.

🎯 What Can EdgeMesh Do?

🖼️ Convert any image to a 3D mesh in seconds
🧠 AI-powered depth estimation using multiple models (MiDaS, DPT, ZoeDepth, Depth-Anything)
🎨 Advanced edge detection with customizable parameters
🔧 Real-time 3D preview with interactive viewport
📁 Export to standard formats (.obj, .stl) for 3D printing or modeling software
🎮 SpaceMouse support for professional 3D navigation

✨ Key Features

🖼️ Image Processing

Smart Background Removal - Automatically detects and removes backgrounds based on corner colors
Edge Detection - Advanced Canny edge detection with adjustable thresholds
Multiple Smoothing Methods - Gaussian, Bilateral, Median, and Anisotropic Diffusion
Grayscale Conversion - Process images in grayscale for enhanced control
Dynamic Depth Adjustment - Create front-half meshes perfect for 3D printing

🧠 Depth Estimation Models

Choose from multiple state-of-the-art models:

MiDaS (Small & Large variants) - Fast and reliable
DPT (Large & Hybrid) - High accuracy
ZoeDepth (K, N, NK, N-indoor) - Specialized for different scenarios
Depth-Anything - Latest cutting-edge models

🎨 3D Mesh Generation

Depth-to-3D Conversion - Direct conversion from depth maps
Edge-based Mesh Creation - Generate meshes from detected edges
Surface Partitioning - Intelligent surface segmentation
Extrusion Projection - Complex 3D reconstructions from contours
Real-time Preview - See your mesh as it's created

🛠️ Professional Tools

3D Text Generator - Create 3D text meshes using the included MeshTools
Mesh Manipulation - Scale, rotate, and transform meshes
Custom Coloring - Apply gradients and color transitions
Measurement Grid - Visualize dimensions and proportions
Batch Processing - Process multiple images efficiently

🚀 Quick Start

📋 Prerequisites

Python 3.12 (⚠️ Not compatible with Python 3.13+)
Windows (primary support) or Linux/WSL
4GB+ RAM recommended
NVIDIA GPU (optional, for faster processing)

🛠️ Installation

Option 1: Automated Installation (Recommended) 🎯

# Clone the repository
git clone https://github.com/lelandg/EdgeMesh.git
cd EdgeMesh

# Run the automated installer
python install_requirements.py

Option 2: Manual Installation 📦

# Clone the repository
git clone https://github.com/lelandg/EdgeMesh.git
cd EdgeMesh

# Update pip
python -m pip install --upgrade pip

# Install requirements
pip install -r requirements.txt

# For Open3D compatibility issues
python install_open3d.py

▶️ Running EdgeMesh

Windows 🪟

python edge_mesh.py
# or simply:
run.bat

Linux/WSL 🐧

python3 edge_mesh.py

📖 How to Use

Step 1: Load an Image 📸

Click "Load Image" and select any image file (PNG, JPG, etc.)

Step 2: Configure Settings ⚙️

Depth Model: Choose your preferred AI model
Smoothing: Select smoothing method and intensity
Edge Detection: Adjust thresholds for edge sensitivity
Resolution: Set mesh resolution (higher = more detail)

Step 3: Process the Image 🔄

Click "Generate 3D Mesh" to create your 3D model

Step 4: Preview and Export 💾

Use the 3D viewport to inspect your mesh
Mouse controls: Rotate, zoom, pan
Export as .obj or .stl for 3D printing or further editing

Pro Tips 💡

Enable "Dynamic Depth" for 3D printing-friendly meshes
Use "Remove Background" for cleaner results
Try different depth models for various image types
Indoor scenes work best with ZoeDepth N-indoor model

🏗️ Architecture

EdgeMesh follows a modular pipeline architecture:

📷 Image Input
    ↓
🎨 Preprocessing (Edge Detection, Smoothing)
    ↓
🧠 Depth Estimation (AI Models)
    ↓
🔨 Mesh Generation (Depth-to-3D or Edge-based)
    ↓
✨ 3D Viewport (Interactive Preview)
    ↓
💾 Export (.obj, .stl)

For detailed architecture information, see the Code Map.

🎯 Examples

Creating a 3D Portrait

Load a portrait photo
Select MiDaS Large model
Enable Dynamic Depth
Generate and export for 3D printing

Architectural Visualization

Load building photograph
Use DPT Hybrid for accuracy
Enable edge detection for sharp details
Export to modeling software

3D Text Creation

python MeshTools/text_3d.py --text "Hello 3D" --height 100 --depth 20

🔧 Troubleshooting

Python Version Compatibility ⚠️

Important: Open3D 0.19.0 is not compatible with Python 3.13+

Solutions:

Use Python 3.12:

# Check your version
python --version

# If 3.13+, install Python 3.12

Create a virtual environment:

python3.12 -m venv venv
source venv/bin/activate  # Linux/Mac
venv\Scripts\activate      # Windows

Run compatibility checker:
```
python install_open3d.py
```

Common Issues

Issue: ImportError with Open3D Solution: Reinstall with specific version

pip install open3d==0.19.0 --force-reinstall --no-cache-dir

Issue: CUDA/GPU not detected Solution: Install PyTorch with CUDA support

pip install torch torchvision --index-url https://download.pytorch.org/whl/cu118

📚 Documentation

📖 Code Map - Complete project structure and navigation
🔄 3D Mesh Creation Flow - Detailed processing pipeline
📝 Version History - Release notes and changes
🛠️ API Reference - Detailed class and method documentation

🤝 Contributing

We welcome contributions! Here's how you can help:

🍴 Fork the repository
🌿 Create a feature branch (git checkout -b feature/AmazingFeature)
💻 Make your changes
✅ Test thoroughly
📝 Commit (git commit -m 'Add AmazingFeature')
📤 Push (git push origin feature/AmazingFeature)
🎯 Open a Pull Request

Development Setup

# Clone with submodules
git clone --recursive https://github.com/lelandg/EdgeMesh.git

# Install in development mode
pip install -e .

# Run tests (when available)
python -m pytest

🌐 Related Projects

🎨 ImageAI by Leland Green

For AI-powered image generation that can serve as input for 3D mesh creation, check out ImageAI - a comprehensive tool supporting multiple AI providers including:

Google Gemini
OpenAI DALL-E
Stability AI
Local Stable Diffusion

Perfect for creating unique images to convert into 3D meshes! 🚀

👨‍💻 Author

Leland Green

🌐 Website: LelandGreen.com
📧 Email: contact@lelandgreen.com
💬 Discord: The Intersection of Art and AI
🐙 GitHub: @lelandg

Also creator of ImageAI - Advanced AI Image & Video Generation Tool

📄 License

This project is proprietary software. Please see the LICENSE file for details.

🙏 Acknowledgments

Open3D Team - For the excellent 3D processing library
PyTorch Team - For deep learning framework
MiDaS Contributors - For depth estimation models
Qt/PyQt6 - For the robust GUI framework
Community Contributors - For feedback and improvements

🚀 Roadmap

Coming Soon 🔜

🤖 More AI models (SAM integration)
📱 Mobile companion app
⚡ GPU acceleration optimizations
🔧 Plugin system for custom processors

Future Vision 🔮

Real-time 3D reconstruction from webcam
VR/AR export capabilities
Cloud processing for large batches
AI-assisted mesh refinement

💖 Support

If you find EdgeMesh useful, please:

⭐ Star the repository
🐛 Report bugs via Issues
💡 Suggest features
📣 Share with others
🤝 Join our Discord Community

Made with ❤️ by Leland Green

🔝 Back to Top

Name		Name	Last commit message	Last commit date
Latest commit History 67 Commits
.claude		.claude
Images		Images
MeshTools @ fbe80ea		MeshTools @ fbe80ea
docs		docs
logs		logs
.gitignore		.gitignore
.gitmodules		.gitmodules
CLAUDE.md		CLAUDE.md
Doxyfile		Doxyfile
EdgeMesh.ico		EdgeMesh.ico
LICENSE		LICENSE
Open3D Help.txt		Open3D Help.txt
README.md		README.md
_version.py		_version.py
build.bat		build.bat
build.sh		build.sh
color_transition_gradient_generator.py		color_transition_gradient_generator.py
depth_anything.py		depth_anything.py
depth_based3d_reconstruction.py		depth_based3d_reconstruction.py
depth_cue_estimator_util.py		depth_cue_estimator_util.py
depth_to_3d.py		depth_to_3d.py
edge_clustering_analyzer.py		edge_clustering_analyzer.py
edge_detection.py		edge_detection.py
edge_mesh.py		edge_mesh.py
edge_mesh.yml		edge_mesh.yml
file_tools.py		file_tools.py
flow_layout.py		flow_layout.py
image_processor.py		image_processor.py
install_open3d.py		install_open3d.py
install_requirements.py		install_requirements.py
log_utils.py		log_utils.py
mesh_generator.py		mesh_generator.py
qt_extensions.py		qt_extensions.py
requirements.txt		requirements.txt
run.bat		run.bat
run.sh		run.sh
shape_analyzer.py		shape_analyzer.py
smoothing_depth_map_utils.py		smoothing_depth_map_utils.py
space_mouse_controller.py		space_mouse_controller.py
space_mouse_event_handler.py		space_mouse_event_handler.py
spacemouse_demo.py		spacemouse_demo.py
spinner.py		spinner.py
surface_partitioning.py		surface_partitioning.py
torch_test.py		torch_test.py
torch_utils.py		torch_utils.py
update_tasks.py		update_tasks.py

License

lelandg/EdgeMesh

Folders and files

Latest commit

History

Repository files navigation