OCR: how to automate OCR of text + math + figures?

Any Questions - Post Here

OCR: how to automate OCR of text + math + figures?

Postby Nag » Fri Oct 27, 2006 10:16 am

Hello:

I have several boxes of papers (with text, math and figures) from pre-pdf era. Would like to scan and then use OCR software to convert to searchable pdf files.

When I tried with 300 dpi scan and ABBY software, it become a mess.

Questions.

1. Should I use higher dpi?

2. Are there any settings, setup files etc for ABBY (or other OCR software) that will recognize text and leave math and figures as they are? This has to be done automatically ... there is way too much stuff to do manually.

3. I am open to suggestions for combinations that work: scanner + OCR software.

Appreciate input from people with experience.

Best
Nag
Random avatar
Nag
 
Posts: 58
Joined: Fri Nov 25, 2005 5:30 pm

Postby Nag » Sun Oct 29, 2006 8:24 am

No response so far!!
An extensive search in google didn't help.

Any ideas, suggestions?
Random avatar
Nag
 
Posts: 58
Joined: Fri Nov 25, 2005 5:30 pm

Postby aihe » Mon Oct 30, 2006 7:51 am

ABBYY Finererader from memory did have a feature where you could select Train User Pattern. I don't know how accurate this was as most of my scanning was text. I rarely had to scan these special characters but I do remember normal OCR would not recognise most of them. Even changing contrast, dpi etc would not work for me. My only solution was to get a good an image as possible and save it as pic file, but as you say, this won't produce a searchable PDF file.
Password:
Code: Select all
www.read.forumsplace.com

Or
Code: Select all
http://read.freeforum.ca
Random avatar
aihe
Founder
Founder
 
Posts: 1603
Joined: Sun Aug 28, 2005 9:04 am


Return to Question & Answers

Who is online

Users browsing this forum: No registered users and 0 guests

cron
Hosted by Freeforum.ca, get your free forum now! TOS | Support Forums | Report a violation
MultiForums powered by echoPHP phpBB MultiForums