AutoInput Query UI for google vision OCR??

Hi,
I'm having a ton of fun with AutoInput. However, I'm having trouble "grabbing" text via AutoInput Ui Query that is displayed in a camera app called "Skaneet". Skaneet (from play store) is an OCR using, as far as I can tell, google vision. You can pass over text with your phone's camera and it will use OCR to turn that real world text into strings. Edit: as text is recognized, it pops up on the app's screen. I'm trying to scan a list of 30 serial numbers, and then output that array of strings into a text file. My initial idea was to query the app every 250ms to capture all text recognized 4 times per second until the app is closed. I'd then put all of that into a text file for later filtering/regex work. Any ideas on how to do this? Much appreciated!

 

Joao! I'm a bit starstruck. Your apps rock man! Thanks a bunch for responding! I'm going to try the OCR in autotools using the method you linked to. Thank you for that. Here's what I wrote to the dev of Skaneet:
Here's what I'd like to see:
You know when you open skaneet and scan, say, a sheet of paper and all of those white words pop up on the screen to show what the OCR is currently recognizing? I want a log of all of the white words that pop up during the time the app is open. When that app is closed, the log is saved as a file to my phone. I can then go through and sort for certain words/numbers after the fact.
Can you think of any simple way to do that with your existing app? Those white words aren't notifications are they? Are they strings in Html? One thought I had was to screen cap (jpeg) the app every 250 ms so that for a 5 second run time of your app, I'd have ~20 jpegs that I could then do OCR on to see what the white words were on each jpeg, thus capturing all of them. Obviously this is a kind of silly way of doing it.
Ultimately, I'm just trying to hack something together so that I can walk up to a shipment of stuff at my work, turn on my camera, pass it over all of the serial numbers of a stack of products/boxes, then close the camera. It would then save a file of all recognized text during the time the camera was on. Does that make sense? Do you have a Patreon account? Thanks again!

 

Thanks Joao! Patreon comment was for other dev (his app was free so wanted to buy him a sandwich for his help). Do you ever do contract work? Thanks!