1
votes

Tesseract OCR is running on iOS 8.x, on an iPhone 5. It typically does well on the first image (from camera), but subsequently produces garbage (as the "recognized text").

I presume there is some kind of internal state in the Tesseract library that needs to be cleared before each re-run (i.e., same app instance, but for all OCR runs after the first). Does anybody know off-hand what, if any, special steps should be taken before a re-run of Tesseract?

Thanks.

1

1 Answers

0
votes