OCR and apple VisionKit ? any advice

x-code
1

I’ve got tesseract OCR running fairly happily on OSX (thanks to @kylemcdonald ofxTesseract addon and the version ofxTesseract3 from @watabOshi GitHub - watab0shi/ofxTesseract3: Addon for using Tesseract OCR library v3.0
but - I am running it against video I’m shooting out of train windows and it’s not really suitable - I’m getting interesting results using the inbuilt VisionKit text recognition in iPhone but am out of my comfort zone on how to get visionKit working out of oF - does anyone know of any existing bridges/addons or have any advice on how to loop to the visionKit routines from oF in a simple manner?

(screenshot of what I am getting with Tesseract - which I really like - just need some more accuracy and less noise…)

Screenshot 2023-05-06 at 18.35.48
Screenshot 2023-05-06 at 18.35.483600×2338 338 KB

2

Hi @danb,
I stumbled against a similar issue some time back as I wanted to use the vision framework in iOS and openframeworks. My goal at the time was to detect faces but recognizing text in images should be fairly similar.

I made an addon, that should be a pretty good starting point:
https://github.com/pierdr/ofxiOSFaceTracking

3

@pierdr thanks for the link and the addon - I will dig through and see what I can make work :slight_smile: