I do not have any recognizer yet. And, about the ground truth I am still yet to build. I plan to setup ground truth per line. I have just used one command, "ocropus page example.jpg". What would be the best suggestion to start the training from scratch?
I'm not very much familiar with all OCR training, so please bare with me. Thanks again, Moiz On May 14, 2:51 pm, Tom <[email protected]> wrote: > On May 14, 8:50 pm, Moiz <[email protected]> wrote: > > > Thanks alot. That worked! > > I've some more question. I am using ocropus-pages command and I see > > pretty nice output on terminal. Does this command store this output in > > any file in hOCR/HTML form? > > You just redirect the output into a file. All the other noise is on > stderr. > > > I also need to train the engine for better > > output, is there any place I can find step by step training guide? > > That doesn't exist yet. There are currently several different > recognizers, and how you do training for each of them is somewhat > different. The recognizer that we actually want to support long term > doesn't have some important tweaks that the old recognizer has, so > I'll write up training after those have been added. > > How you train also depends on whether you already have a basic > recognizer and just want to improve performance or whether you're > training for a new script. And it depends on whether your ground > truth is per character, per word, per line, or per page, and how > accurate your ground truth is. > > Tom > > -- > You received this message because you are subscribed to the Google Groups > "ocropus" 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 > athttp://groups.google.com/group/ocropus?hl=en. -- You received this message because you are subscribed to the Google Groups "ocropus" 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/ocropus?hl=en.
