AI Potluck
Back to Gap Map Infrastructure / Classic ML & computer vision

OpenCV

OpenCV
open source / Overall score: 3.4

Open Source Computer Vision Library: C++ image and video processing, feature detection, camera calibration, stereo, stitching and object detection routines with Python, Java and JavaScript bindings, plus a dnn module that runs ONNX, TensorFlow and TFLite networks. The opencv-python wheels on PyPI are built from a separate packaging repository. Maintained by the non-profit Open Source Vision Foundation.

Openness

5 medium confidence
5.0
license
Apache-2.0(OSI
source
public(github.com/opencv/opencv)
core features withheld
no — maintained by a non-profit foundation

OpenCV is Apache-2.0 and the library builds from its public repository, with extra modules in the equally public opencv_contrib. The only algorithms missing from the PyPI wheels are patented ones such as SURF, left out because they cannot be redistributed, not held back for a paid tier.

Adoption

5 high confidence
5.0

Measured on monthly PyPI downloads of opencv-python; the contrib and headless wheels are separate packages and not counted.

Capability

1 medium confidence
1.0

OpenCV supplies the building blocks of computer vision - filtering, geometry, features, video and a fast engine for running networks trained in other frameworks - rather than training a predictor itself. Its classical learners now sit in the separate contrib repository, so the main library fits no model of its own.

Verified 2026-09-26