https://github.com/andre-silva-14/heightcraft
An high-performance, heightmap generation and AI upscaling tool
https://github.com/andre-silva-14/heightcraft
3d 3d-reconstruction 3dmesh automation computational-geometry digital-elevation-model game-dev game-development gis heightmap lidar mesh photogrammetry point-cloud python sea-level terrain-generation
Last synced: 29 days ago
JSON representation
An high-performance, heightmap generation and AI upscaling tool
- Host: GitHub
- URL: https://github.com/andre-silva-14/heightcraft
- Owner: andre-silva-14
- License: mit
- Created: 2024-11-20T21:27:39.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2025-12-04T08:08:50.000Z (2 months ago)
- Last Synced: 2025-12-05T05:10:29.682Z (2 months ago)
- Topics: 3d, 3d-reconstruction, 3dmesh, automation, computational-geometry, digital-elevation-model, game-dev, game-development, gis, heightmap, lidar, mesh, photogrammetry, point-cloud, python, sea-level, terrain-generation
- Language: Python
- Homepage: https://andre-silva-14.github.io/heightcraft/
- Size: 7.29 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
**Heightmap Generation & AI Upscaling Tool**
[](LICENSE)
[](https://www.python.org/)
[](./docs)
[](./tests)
**Heightcraft** is a high-performance CLI tool designed for **Game Engineers**, **GIS Professionals**, and **Researchers**. It automates the pipeline of converting 3D data into high-precision heightmaps, featuring accessible **AI Upscaling**.
---
## 🚀 Key Features
- **🤖 Heightmap AI Upscaling**: Transform low-res inputs into crisp, hi-res/high-detail heightmaps. Also supporting bit depth increase (8-bit → 16/32-bit) with AI hallucination.
- **📡 LiDAR Support**: Stream process massive `.las` and `.laz` point clouds into Digital Elevation Models (DEMs).
- **🏔️ Mesh to Heightmap**: Bake `.gltf`, `.glb`, `.obj`, `.stl`, and `.ply` meshes into heightmaps with automated chunking.
- **🌊 Feature Masks**: Optionally export water masks with sea level thresholds and texture masks for slope and curvature.
- **🎯 High Precision**: Native support for **32-bit Float TIFF**, **16-bit PNG**, and **RAW** formats.
- **⚡ Performance**: GPU acceleration and memory-efficient streaming for gigabyte-scale datasets.
## 📖 Documentation
Full documentation is available in the [Docs](https://andre-silva-14.github.io/heightcraft/).
- [**Getting Started**](https://andre-silva-14.github.io/heightcraft/docs/getting-started)
- [**LiDAR Processing**](https://andre-silva-14.github.io/heightcraft/docs/guides/lidar-processing)
- [**Mesh Processing**](https://andre-silva-14.github.io/heightcraft/docs/guides/mesh-processing)
- [**AI Upscaling Deep Dive**](https://andre-silva-14.github.io/heightcraft/docs/guides/ai-upscaling)
## 📦 Installation
```bash
pipx install heightcraft
```
## ⚡ Quick Start
**Upscale an image (8-bit → 16-bit and 3x upscale):**
```bash
heightcraft input.png --upscale --upscale-factor 3 --bit_depth 16 --output_path high_res.png
```
**Process LiDAR data into an high-res 32-bit heightmap:**
```bash
heightcraft scan.laz --bit_depth 32 --chunk_size 600000 --max_resolution 24576
```
**Convert a 3D Mesh to an upscaled heightmap:**
```bash
heightcraft terrain.obj --large-model --chunk_size 200000 --max_resolution 8192 --upscale
```
## Development
To install development dependencies:
```bash
pipx install uv
uv sync --group dev
uv run main.py
```
## 🧪 Testing
Run the test suite:
```bash
pytest tests
```