moviefingerprint analyzes a video stream and returns an image that represents the movie's 'fingerprint'. This fingerprint image is a unique type of image averaging that maintains the most common ambient colors and image patterns.
Detailed description and examples here
- Python 3.6 (3.5 should work as well)
- numpy - install with pip or conda
- OpenCV v3.2 - OpenCV is installed by downloading the '.whl' file here (make sure to grab the correct version), navigate to the file's directory, and run 'pip install opencvfilename'
import pymoviefingerprint
mfp = pymoviefingerprint.MovieFingerprint(r'moviepath', 'movietitle') # raw string (r') is helpful with paths
print(mfp.movie_title) # returns the name of the movie
print(mfp.total_frames) # returns the total number of frames in the movie
mfp.make_fingerprint() # Generates fingerprint image (does not save image)
mfp.write_fingerprint_image() # Saves image to /images/ folder
mfp.get_matching_image() # Generates closest matching movie frame to fingerprint image (does not save image)
mfp.write_matching_image() # Saves image to /images/ folder