Re: [tesseract-ocr] Re: image_to_string OSD hell

2024-02-14 Thread dev 313153
Thanks a lot for your answer ! After playing around, the issue is that apparently both whitelist and blacklist aren't supported in this scenario and make tesseract return nothing, but i don't really understand why because it works find in another scenario (for whole picture recognition, before

Re: [tesseract-ocr] Re: image_to_string OSD hell

2024-02-14 Thread Tom Morris
On Wednesday, February 14, 2024 at 5:50:45 AM UTC-5 dev 313153 wrote: the issue is that apparently both whitelist and blacklist aren't supported in this scenario and make tesseract return nothing, Either that or you're specifying them wrong. You appear to have two different config settings, bu

[tesseract-ocr] Re: Traineddata files

2024-02-14 Thread Tom Morris
On Tuesday, February 13, 2024 at 12:51:35 AM UTC-5 argo...@gmail.com wrote: What if there is no traineddata files for a language ? How do I start building a trained data file for the breton language ? Searching the archives / group for "training from scratch" should turn up lots of previous di

Re: [tesseract-ocr] Re: I need help to develop image to text extraction

2024-02-14 Thread AROMAL
then go for the deep learning models since you have a dataset it will be easy and less complex for the word level text extraction task :) On Tuesday 13 February 2024 at 11:10:53 UTC+5:30 santhi...@gmail.com wrote: > Word level extraction only > > On Tuesday 13 February 2024 at 11:10:03 UTC+5:30