text scanning

[ERICBRIAN2002]

I have a umax astra 2000u. I have a paper some one wrote with mistakes on it and so i thought instead of rewriting the whole paper i could scan it and make the corrections and print it out. what software do use to do this and how thanks again.

[FaRuvius]

The scanning software you are looking for is something called Optical Character Recognition, or OCR.

I do not know of any freeware/shareware programs that do OCR.

There are several commercial apps available for OS9.

http://www.versiontracker.com/mp/new...er+recognition

There is one commercial app available for OS X:

OmniPage Pro for OS X

Hope this helps,

FaRuvius

[edX]

omnipage ltd ed should have come with your scanner. will probably only work in os9 but might run in classic

[ERICBRIAN2002]

I found the software i needed but when i scanned it and it showed the text they where all numbers and letters didnt even make sence no where near the paper? why is this? thanks again for the software tho you guys and girls are great.

[genghiscohen]

I use OmniPage Pro under Classic for doing OCR, and it works just fine.
One mistake that I made the first time I used it - I put the page I was scanning in upside down, and got back total gibberish.

[edX]

ocr scanning is as much an art as it is a science. 1st, if this paper has any staples or clips, remove them. you want to get that paper as flat as possible with only one sheet being in the scanner at a time. second, make sure that it is as perfectly aligned as possible. any misalignment will make a difference. third, the original font makes a difference. in general fonts with serifs are easier to scan just as they are easier to read.

s e r i f s

rest of post has no serifs

lastly, i am working under assumption this is a normal wp doc with a single column. scanning multiple column docs is much trickier but can still be done.

[ERICBRIAN2002]

It was a paper that was handed out. I dont know what type of font it is. When i scanned it it can out clear but then when they put it in text its a whole bunch of numbers and letter.i dont get the scan comes out good but then when the text open its messed up.

[chemistry_geek]

OCR software is dependent on the resolution of the scan. A number of years ago I purchased an HP scanner (that I now no longer own) that came with OCR software.

The requirements at that time for GOOD optical character recognition were:

1. The files MUST be in TIFF format and be black and white.

2. The scans had to be at least 150 dots per inch. The software was setup to recognize the Courier 12-point font. It would not recognize characters less than 12-point.

3. Staples and creases in the paper interferred.

4. The paper had to be aligned properly on the glass (flatbed scanner).

Hope this helps.

When I used it, it did a pretty good job, but it ALWAYS required some corrections and was NOT 100%. I would say it was at least 85% to 90% accurate. It often mistaken an "e" for an "o" and vice versa.

One more thing, I only used OCR TWICE in my life, both times were for academic purposes - mass data incorporation for research papers. Aside from this, I would never have a use for OCR.