This is very cool. I'm curious about how you process the video feed; Do you stitch it back together as one long image and then perform OCR? Or just look at eve

This is very cool. I'm curious about how you process the video feed; Do you stitch it back together as one long image and then perform OCR? Or just look at every N frames? I've been looking for an open source library that helps with taking a scrolling video and then stitching it into an image but I haven't had much luck.