Description
Video Processing SciKit BETA
Video processing algorithms, including I/O, quality metrics, temporal filtering, motion/object detection, motion estimation...
This is intended as a companion to scikit-image, containing all the algorithms which deal with video. There is a certain degree of overlap between image and video algorithms, for example a PSNR quality metric could be applied to pairs of images or pairs of video frames just as well. However, other algorithms are video-specific, for example a temporal denoise. This is the future home of the video-specific algorithms, as well as some of the algorithms which are not strictly video specific but are usually seen in a video context.
This also has some overlap with OpenCV. Roughly, the algorithms implemented here would be easier to hack on, and more research-oriented. Rather than building on top of a C/C++ framework, this will stay Python all the way, using whichever combinaiton of Numba/Theano/etc seems best for performance. This should add flexibility and better future ability to use GPU compute.
The project milestones are roughly:
scikit-video alternatives and similar packages
Based on the "Video" category.
Alternatively, view scikit-video alternatives based on common mentions on social networks and blogs.
-
vidgear
A High-performance cross-platform Video Processing Python framework powerpacked with unique trailblazing features :fire:
WorkOS - The modern identity platform for B2B SaaS
* Code Quality Rankings and insights are calculated and provided by Lumnify.
They vary from L1 to L5 with "L5" being the highest.
Do you think we are missing an alternative of scikit-video or a related project?
README
scikit-video
Video Processing SciKit BETA
Video processing algorithms, including I/O, quality metrics, temporal filtering, motion/object detection, motion estimation...
This is intended as a companion to scikit-image, containing all the algorithms which deal with video. There is a certain degree of overlap between image and video algorithms, for example a PSNR quality metric could be applied to pairs of images or pairs of video frames just as well. However, other algorithms are video-specific, for example a temporal denoise. This is the future home of the video-specific algorithms, as well as some of the algorithms which are not strictly video specific but are usually seen in a video context.
This also has some overlap with OpenCV. Roughly, the algorithms implemented here would be easier to hack on, and more research-oriented. Rather than building on top of a C/C++ framework, this will stay Python all the way, using whichever combinaiton of Numba/Theano/etc seems best for performance. This should add flexibility and better future ability to use GPU compute.
The project milestones are roughly:
- Add skeleton project from scikit-example - DONE
- Add video I/O by wrapping ffmpeg/avconv (similar to kanryu/pipeffmpeg) - DONE
- Add video metrics (from aizvorski/video-quality) - DONE
- More contributions roll in :)