I’m writing a Python script to extract some images from a TV signal. I’ve managed to import the data into a numpy array and found the start point of the image sequence, but each scan (line) of the image is part of a carrier wave. How do I remove this wave to leave just the image data?

If I look at the data in Audacity, an empty part of the signal looks like this:

A part of the signal that contains data looks like this:

In this case, each scan contains two peaks.

I tried using a highpass filter to flatten the wave a bit and to see if the peaks stood out, but it didn’t really help.

I don’t know much about signal processing, so there may be something in scipy that does this out of the box. As far as I can tell, what I need to do it identify the low peak that is the start of the scan, identify peaks in the next 12ms or so (or identify the gradient change at the end of the signal), then move on to the next scan. Can anyone suggest the best way to do this?