Python / C ++ / Java: Is there a library that recognizes characters and tells me where they are in the image?

Question

I looked at PyTesser / Tesseract and, as far as I can tell, it gives me a line only after inputting the image.

However, I would like to get instead the coordinates / bounding box of each recognized character. The string itself may be useful to me later, but right now, that now is what I need.

Assuming PyTesser / Tesseract does not offer me this functionality, is there something freely available (i.e. not Matlab or something like that)?

Answer 1

According to this mailing list thread, you can do this with Tesseract. However, I do not know about the Python shell that will give you access to this data.

Answer 2

Look at a similar question recently posted to the tesseract google group: http://groups.google.com/group/tesseract-ocr/browse_thread/thread/70ad57cb37edabe0/

Answer 3

Perhaps you should try "ocropus":

http://code.google.com/p/ocropus/