aizvorski

scikit-video: Python Video Processing Algorithms

scikit-video is a Python library for video I/O and processing, designed as a video-focused companion to scikit-image. It includes quality metrics, temporal filtering, and motion analysis routines for research and experimentation.

View on GitHub Compare
Stars
153
Forks
23
Stars, last 30 days
+0
Last commit
6.7 years ago
Contributors
2

Our take

Inactiveno commits in over a year

Health score

20/100

How it is scored
Activity
19
Community
18
Issues
21
Pull requests
20

+0 stars in the last 30 days

  • Fewer contributors than 86% of the projects we track

scikit-video covers useful research-oriented video tasks, but its inactive status and reliance on a highly concentrated contributor base make it a cautious choice for new production dependencies.

Good fit if

  • You need Python routines for video I/O, metrics, temporal filtering, or motion analysis.
  • You are comfortable evaluating beta software and maintaining or adapting a research-oriented library.
  • You can work with a project whose recent development and issue handling have stopped.

Look elsewhere if

  • You need active maintenance, recent releases, or responsive issue and pull-request handling.
  • Your team cannot absorb the risk of most contributions coming from one person.
  • You need a clearly asserted license before adopting the code; the available repository metadata reports no asserted license.
All health signals
Last commit2020-02-28 (6.7 years ago)
Commits, last 90 days0
Releases, last 12 months0
Contributors2 (top contributor: 92% of commits)
Issues closed, last 90 days0
Pull requests merged, last 90 days0
Project age12 years

Checked on 2026-11-06 with the GitHub API.

Project history

Daily GitHub snapshots since Oct 7, 2026. History covers stars and forks; health signals are not tracked historically.

Selected snapshot

Stars

153

Forks

23

Change since Nov 16, 2026: 0 stars·0 forks

Overview

scikit-video is a Python video-processing library intended as a companion to scikit-image. It brings video I/O and video-focused algorithms together, addressing tasks such as quality measurement, temporal denoising, and motion analysis that are not handled as naturally by image-only tools.

The project describes itself as a research-oriented, Python-first alternative for video algorithms, with I/O built around ffmpeg or avconv. Its aim is to make these routines easier to modify than code built on a C/C++ framework. The README labels the project beta.

Key Features

  • Video input and output through ffmpeg or avconv wrappers.
  • Video quality metrics, including metrics carried over from the video-quality project.
  • Temporal filtering, including temporal denoising.
  • Motion and object detection routines.
  • Motion estimation algorithms.
  • A Python-oriented codebase intended to be easier to hack on for research work.

Use Cases

  • Researchers prototyping or modifying video-processing algorithms in Python.
  • Developers adding video I/O and quality measurements to SciPy-oriented workflows.
  • Teams exploring temporal denoising on video rather than processing each frame independently.
  • Experimenters implementing motion or object analysis without starting from a C/C++ framework.

What you need

Detected in the repository

  • Python (from setup.py)

Getting Started

The repository identifies Python as its runtime. Start with the project README for installation instructions and usage details: https://github.com/aizvorski/scikit-video

License in Plain Words

No standard license was detected. Without a license, the default copyright rules apply: you may read the code, but not reuse it. Check the repository or ask the authors.

Alternatives

  • vidgear: VidGear focuses on real-time capture, encoding, and streaming, while scikit-video emphasizes research-oriented video processing, metrics, and analysis.
ProjectLanguageLicenseStarsStatus
scikit-videoPython–153Inactive
vidgearPythonApache-2.03.7kQuiet

Considerations

The repository is marked inactive, with no recent commits or releases, and no issues or pull requests handled in the recent measurement window. The measured contributor base is small and contributions are highly concentrated, so ongoing maintenance may depend on one person. The README calls the software beta and describes some performance approaches as future possibilities rather than established project capabilities. No asserted license is available in the supplied repository metadata, so check the repository and its files before relying on it or redistributing it.

Found this useful?

Share it with someone who would like scikit-video.

Comparisons