Can you get better results by having more than one picture of the same text?

[maxm007]

Hi,

I'm doing feasibility study and looking at OCR components. I want to
grab text from video in real time. So imagine pointing a camera at a
poster. The user focuses on a piece of text and it would return real
time ocr results. When the user is happy that it's accurate he would
accept the result.

Most OCR solutions seem to work with one input image to base their
results on. I imagine a burst of images of the same text could
provide better results. Could you optimize tesseract to work in this
way or would that have to be done by the client code that uses
tesseract library.

Thank you

[Svetlin Nakov]

Hi Max,

In particular situations Tesseract will do the job but generally it has 2
limitations:

1) It doesn't work well for images using less than 300 DPI (e.g. it can not
recognize screenshots at with text 96DPI)

2) You should implement a filter to separate the text from the background
and this is really a challengeable task

Regards,

Svetlin Nakov
Managing Partner
Consulting and Information Technology Agency (CITA)
http://www.citagency.eu

--

You received this message because you are subscribed to the Google Groups
"tesseract-ocr" group.
To post to this group, send email to tesser...@googlegroups.com.
To unsubscribe from this group, send email to
tesseract-oc...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/tesseract-ocr?hl=.