This skill provides a high-fidelity workflow for video shot breakdown by analyzing scene change scores to identify cut points. It leverages ffmpeg and ffprobe to intelligently extract the first and last frames of every shot, offering features like gap analysis for noise reduction and smart handling of edge cases such as rapid montages, fade-ins, and dissolves. It is an essential utility for video editors, data scientists, and developers needing automated, precise frame extraction for archival, visual logging, or training data generation.
主要功能
01Automated gap analysis to recommend optimal detection thresholds
02Adaptive scene detection with user-tunable thresholds
03Extracts first, last, or both frames per individual shot
04Cached analysis scores for instant iteration and threshold adjustments
05253 GitHub stars
06Run-based deduplication to filter aftershocks and transition noise
使用场景
01Generating visual shot logs and storyboards for film production and editing
02Automating the creation of high-quality thumbnails for video content libraries
03Pre-processing video datasets for computer vision and machine learning models