A Key Frame is a location on a video timeline which marks the beginning or end of a smooth transition throughout the fotograms, Key Frame Detector try to look for the most representative and significant frames that can describe the movement or main events in a video using peakutils peak detection functions.

Requirements

• python3
• numpy
• opencv
• peakutils
• matplotlib
• PIL

pip install video-keyframe-detector

After installing the package (see Installation section), you can run the keyframe detector from the terminal:

# After installation, run the detector with your video file:
video-keyframe-detector -s "path/to/your/video.mp4" -d output/ -t 0.3

# Or if running from source:
# git clone https://github.com/joelibaceta/video-keyframe-detector
# cd video-keyframe-detector
# pip install -r requirements.txt
# python3 cli.py -s "videos/acrobacia.mp4" -d output/ -t 0.3

You can also use the detector directly in your Python code:

from KeyFrameDetector.key_frame_detector import keyframeDetection

source_video = "videos/acrobacia.mp4"
output_dir = "output"
threshold = 0.3

keyframeDetection(
    source=source_video,
    dest=output_dir,
    Thres=threshold,
    max_keyframes=5,      
    plotMetrics=True,     
    verbose=True         
)