/CoreOCR

OCR using only CoreML and Vision - completely written in Swift

Primary LanguageSwiftMIT LicenseMIT

CoreOCR

Attempt to create an OCR-App only using CoreML and Vision written in Swift

Remarks:

IOS 11 (and up ) This Demo works in LandscapeRight (HomeButton on right side)

OCR is ONLY trained for Font "INCONSOLATA" (downloadable at Google Fonts)

ToDo:

Better preProcessing

Faster crop-resize-grayscale conversion