kornia
by kornia
Kornia is a Python library, built on PyTorch, that offers image processing, camera geometry and feature-matching tools which can be used inside machine learning code. It is meant for building and training computer vision models and pipelines, especially on a GPU.
- Stars
- 12.2k
- Licence
- Apache-2.0
- Runs on
- Your computer
- Setup
- Moderate
- Language
- Python
- Age
- 2994 days
What you can do with it
- Apply batched image filters, warps and colour adjustments on a GPU as part of a PyTorch pipeline
- Augment images, masks, boxes and keypoints together when training vision models
- Match features between photos and estimate camera geometry using methods such as LoFTR, LightGlue and RANSAC
Make money with it
A freelance developer could offer custom computer vision pipelines, such as image alignment or feature matching, to clients using this library.
An idea, not financial advice. Check the licence before charging for anything built on it.
Ways a business could use it
- Build image-processing steps for product photo checks or visual inspection tools without writing the algorithms from scratch
- Speed up training data preparation for a custom vision model by running augmentations on a GPU
Can you use it commercially?
Free to use, change and sell, including in paid products. Keep the copyright notice.
Not legal advice. Read the project's licence file before building a product on it.
Getting it running
It is installed as a Python package from PyPI (pip install kornia) and needs PyTorch and some Python coding to use.
Follow the official instructions in the project README.
Things to check
- Requires programming knowledge; it is a library, not a finished app
- A GPU is needed for the speed gains described, though it also runs on a CPU
- For simple single-image tasks, OpenCV or torchvision may be faster or simpler
Who's talking about it
Checked 03 Nov 2026. Summary drafted by AI from the project's README.
