# kornia/kornia

🐍 Geometric Computer Vision Library for Spatial AI

Repository: https://github.com/kornia/kornia
Canonical: https://ross.abutalabs.com/products/kornia
Homepage: https://kornia.readthedocs.io
Language: Python
License: Apache-2.0
License Family: permissive
Topics: computer-vision, image-processing, machine-learning, pytorch, deep-learning, neural-network, python, artificial-intelligence, robotics, spatial-ai, hacktoberfest
Last push: 2026-08-26T21:11:19+00:00

## Health v2 (maintenance only)
Score: 86/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 60, longevity 100
- inputs: {"age_days": 2933, "days_push": 7, "days_rel": 106, "gap_med": 150.5, "n_releases_24m": 5}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 11327, forks 1236 (observed 2026-08-28T04:10:46.748399+00:00)

## What it is
Kornia is a differentiable computer vision library built on PyTorch, offering GPU-accelerated image processing, augmentations, and geometric vision algorithms. It includes pre-trained models for tasks like detection, matching, and segmentation, and integrates into deep learning pipelines with auto-differentiation support.

## Use cases
- augment images for training deep learning models
- estimate homographies and camera calibration
- detect edges and apply differentiable filters on GPU
- match image features and stitch panoramas
- run pre-trained vision models like SAM or LoFTR
- compute differentiable geometry losses for training

## When to choose
- you need differentiable image operations inside PyTorch training loops
- you want GPU-accelerated batch image augmentation
- you work on geometric vision like SLAM, calibration, or 3D reconstruction
- you need pre-trained vision models with a PyTorch-native API

## When to avoid
- you need a lightweight CPU-only image processing tool like OpenCV for simple scripts
- you don't use PyTorch and can't take its dependency
- you need production image processing pipelines without deep learning

## Facets
- artifact type: library
- maturity: active
- function: image-processing, computer-vision, machine-learning, deep-learning
- domain: computer-vision, image-processing, deep-learning, machine-learning, robotics
- platform: python, cross-platform
- tags: pytorch, differentiable-programming, data-augmentation, geometric-vision, camera-calibration, image-matching, spatial-ai, gpu

## Member repositories
- kornia/kornia (main) score 86

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:46.748399+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:15:57.167157+00:00, confidence not recorded.
  - readme: https://github.com/kornia/kornia (fetched 2026-08-28T04:10:46.748399+00:00, sha fdf2bb750107)
  - homepage: https://kornia.readthedocs.io (fetched 2026-08-29T08:14:31.635543+00:00, sha 19ce5703c397)
  - registry_pypi: https://pypi.org/pypi/kornia/json (fetched 2026-08-29T08:14:31.638360+00:00, sha 287122d4d413)
- Data as of 2026-08-30T08:39:29.467469+00:00.