OpenCV: detect text from multiple boundaries simultaneously

Question

I\'m currently trying to write a script to detect text in an OBS video stream using Python/OpenCV.

From every n-th frame, I need to dete