I want to attempt to read the text printed on a sheet metal that is moving and stops on the exact same place for 2 seconds and then moves again. I tried to capture some stills manually and recognize text from them and it worked quite well, but the problem is that I never know when the next text will arrive and stop above the camera. I came up with idea as follows: since the text is printed with custom symbol at the end I want to train Tensorflow lite model to capture a still when a symbol is recognized and then do the ocr from the still. What do you think? Do you know any easier way of approaching this? Main problem is that sometimes we have 10 seconds between stops and sometimes there is nothing happening for 30 minutes.