Free OCR: Turn a picture of text into real text without spending a dime
Optical character recognition (OCR) software has been around for a long time. But for the most part, it's still cumbersome and expensive. If you only occasionally need to turn an image of text into the real thing, there's no point in buying it. So here are two simple OCR solutions that won't make you go through a complex installation or pull out your credit card.
[Have a tech question Ask PCWorld Contributing Editor Lincoln Spector. Send your query to answer@pcworld.com.]
If you're unsure about the difference between an image of text and real text, try this experiment: Double-click a word in this paragraph, copy it, and paste it into a word processor. Now try doing the same with a word in the image below. See the difference
You may already have an OCR program. OneNote, the outliner and research organizer that comes with many versions of Microsoft Office, has had OCR capabilities since version 2007. If you don't have Microsoft Office, OneNote is also available as a free download, although you will be required to use or create a Microsoft account.
Using it for OCR is very simple. Just copy and paste the image into a OneNote page. Then right-click the image and select Copy Text. OneNote will OCR-copy any text it finds in the image text to the clipboard.
If you don't want to use OneNote, try the Free OCR website. All you need to do is go to the site, upload the file, and the text will appear in a box on the page. You don't even have to give anyone your email address.
Is it safe The Free OCR Privacy Page promises that "We will not view the files that you upload using the Free OCR Service. Your files are deleted after processing." On the other hand, the site isn't secure--no SSL. I wouldn't use it for anything sensitive.
Both OneNote and Free OCR generally produce accurate text, but you should check it yourself afterward. This is especially true if the image is complex, with a lot of graphics mixed with the text.
Fonts and resolution make a difference as well. If the text image is too small, or if it uses a script or unusual font, even the best OCR programs may fail with it.