I want to scan an obituary into the database but instead of image, I'm trying to convert to text using OCR and keep getting jumbled words. Has anyone done this with success? Darci

It all depends on the quality of the scanner, the document, and the OCR program. If you are using the OCR program that came with your scanner, it probably doesn't work very well. You can get a better one for around $100, or a great one for about $1000. AABBY Fine Reader, Omni Page, and Text Bridge Pro are three that I have used and all work pretty well. For me and my arthritic hands, any are an improvement over typing. Norma

What does this have to do with Family Origins? I am not scolding, it's just that Family Origins has no facility for OCR. That has to be done with a dedicated OCR program. I think that some will read characters from a picture file, but the one I have has to have control of the scanner as it works. Once you get the text into a file, clean it up with a good word processor, THEN copy it into a Family Origins note or source. The type of document has quite a bit to do with the accuracy of the OCR too. I have found some OLD newspaper clippings are so yellowed that it is faster, even for me, to type it than it is to try correcting the OCR generated file. A good, clean, well spaced copy works the best. If the copy is small, old, crowded and faded, forgedaboudit. Good Luck, Alfred ========== ----- Original Message ----- From: "Darci" <djohn@netnet.net> > I want to scan an obituary into the database but instead of image, I'm > trying to convert to text using OCR and keep getting jumbled words. Has > anyone done this with success? Darci > >

Darci, Many times, using the same computer and scanner but wildly different quality copies of Obits. And the results vary as wildly as the quality of the obit originals. Sometimes you can get nearly every word recognized..... and sometimes it's impossible to get more than about half recognized. General "rules" I follow Scan at 600 dpi Try scanning in "line art mode" and in "grayscale". Sometimes grayscale gives better results, sometimes not. If working from newsprint (what I usually have) and it's yellowed a lot, it sometimes helps to make a Xerox copy of the obit, and then scan that, instead of scanning the original newsprint. If the original shows some print on the "back" side bleeding through, this can really mess up your OCR. Try scanning with a black sheet of paper backing up your original. Darci wrote: > > I want to scan an obituary into the database but instead of image, I'm > trying to convert to text using OCR and keep getting jumbled words. Has > anyone done this with success? Darci > -- Charlie Hoffpauir http://freepages.genealogy.rootsweb.com/~charlieh/

At 06:49 PM 5/31/01, Darci wrote: >I want to scan an obituary into the database but instead of image, I'm >trying to convert to text using OCR and keep getting jumbled words. Has >anyone done this with success? Darci Darci, If you have a copy machine (don't all genealogists?) make a copy of the obit. OCR scans are much, much better then news print. --Derick