Robust Redundant Video Grader
In today’s world, there are hardly any methods to detect two similar videos. If we need to check if the two videos
are similar or not we manually need to browse the entire length of the videos and check if they are similar or not.Large video
repositories contains a huge amount of videos stored and large amount of memory is wasted when they contain duplicate
videos. So there is a need to develop a technique by which we can identify two duplicate videos. Duplicate video detection is
also required in case of some editing or mixing is done in the video to identify the original video present in the edited video.
We present an efficient and accurate method for duplicate video detection in a large database using video fingerprinting
technique. Video fingerprint is nothing but the unique frames generated of the video. We create the frames from the videos
and compare those frames for duplication. Frames are compared on pixel by pixel method and result is generated. The
technique helps in efficient retrieval of duplicate videos and saves the time of manual browsing and checking the videos for
similarity. We are creating a desktop based application designed in a .NET framework. By using this technique we can find
two duplicate videos and also can find original video if in case any editing or mixing is done with the video.