We need an application that will determine duplicate videos in the media library (about 50.000 videos).
Requirements for the application.
We want the system to detect duplicate videos even if;
-same video is with another encoding (.mp4, .avi, .mpeg ...)
-same video has less than 20 seconds of intro or outro added.
-same video has added logotype on all frames.
-same video is cropped.
Application should run under linux OS.
1. One time it should run against the whole media library, and save duplicate video groups and results of generated fingerprints into the database.
2. Each time a new video is added into library it should be compared against existing database and give the list of possible duplicates.
Free to choose. Just have in mind that library is big and it should run fast.