PySceneDetect is a suite of tools for identifying cuts and transitions in video files using content, threshold, and histogram detection algorithms. It functions as a scene detector, frame extractor, statistics analyzer, metadata exporter, and video scene splitter.
The project identifies scene boundaries and can divide video files into smaller clips using external processing tools. It allows for the extraction of representative image frames from detected changes and the export of scene lists into industry-standard formats such as EDL, FCP, HTML, OTIO, and CSV.
The toolset includes capabilities for per-frame metric tracking and visual pattern analysis to determine optimal detection thresholds. It provides an interface to ingest video content from local files, hardware cameras, or data pipes.
