I have two video signals. One encoded in MPEG format and the other in H.263 format. These signals are represented in .txt files with the frame sizes in an array.
I need to calculate the cross-correlation without losing or adding data, even though they have different sizes.
The file with MPEG frames has 89000 frames and H.263 has 17000 frames.
Any suggestions on how to do this in python or matlab?
What I managed to do was use numpy.correlate with the 'full' parameter which calculates the cross-correlation for all possible offsets (lags) of the time series
correlation = np.correlate(mpeg_format['FrameSize'], h_format['FrameSize'], mode='full')
Attachments: