On 20 May 2013 15:08, <[email protected]> wrote: > El dilluns 20 de maig de 2013 13:29:38 UTC, jimregan va escriure: >> There was a version of Tesseract that did this, using OpenFST, in one of >> the Android trees (I think the original AOSP tree), but you'd have to dig >> through old revisions to find it. > > Looks like it is this one: > > https://android.googlesource.com/platform/external/tesseract/+/d544c9231465999ad600ec13614b4d69d351798d/ > > The date is 3 years and 10 months ago.
That's just the date it was committed to AOSP - it's older than that! > Have any substantial improvements in > OCR quality been made in the main branch since then ? e.g. If I just work > with this branch will I get much worse results than using HEAD ? Yes, and yes. The language model parts shouldn't be all _that_ different, and that's all you're really concerned with, but there are a number of other differences in the code - it includes text detection, for one, and I think it has some neural network code that may or may not be related to the nn code in Cube. -- <Sefam> Are any of the mentors around? <jimregan> yes, they're the ones trolling you -- -- You received this message because you are subscribed to the Google Groups "tesseract-ocr" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/tesseract-ocr?hl=en --- You received this message because you are subscribed to the Google Groups "tesseract-ocr" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/groups/opt_out.
