Prizmo is a pretty amazing iPhone app for OCR and more
There are a few apps that are great demos for the iPhone. Pulse News is a super RSS feed reader, iMovie always drops a few jaws, and Siri is a pretty amazing way to talk to your iPhone and get intelligent search results.
You can add Prizmo to the list of great demo apps, and it's going to be a very useful app for a lot of people. Prizmo combines OCR (Optical Character Recognition), text to speech, and translation all in one place.
Prizmo, which sells for US $9.99, can scan business cards, books, white boards, or even receipts. It can output the text to the clipboard or email, and can even take a receipt and put it into a spreadsheet format.
If that isn't enough, Prizmo can read the text to you with the addition of an optional text to speech module. There are 35 voices you can choose from, in ten languages. Each module is $2.99.
%Gallery-100014%
Most of the program features do not need an internet connection. Scanning and OCR are built into the program. For translation, which uses Google Translate, you'll need internet connectivity.
Does it work? Yes, and quite well. For most OCR tasks, the accuracy is very high. Like most OCR apps, you may have to fix a letter or some errant mistake, but my results were very good. You have to have a good image to start with, well lit and in focus. The iPhone camera will take care of the focus, and the app provides super-imposed grid lines to help you hold the camera straight.
After you take an image or import it from your camera roll, you can crop and enhance the image to assure a good capture. You can also rotate the image, because the OCR software wants to see the text oriented correctly. In addition, you can correct any geometric distortion in the image to make the OCR go more smoothly and accurately. There is also an option to white balance the image before the OCR step.
The OCR process is very fast, and in a couple of seconds you have the text. From that point, you can copy the text, email it, or use Dropbox or CloudApp.
A couple of things I didn't like. The app appears to have no easy way to capture the addresses on a business card. The OCR picks it up, but the contact template doesn't include an address field. You can create one, but you'd have to re-type everything into that field. It's a major shortcoming if your main use of this program is for scanning business cards. I also thought the built-in help was very superficial. It connects to the developer's website, but the text is small and the information is sparse. If you visit the Prizmo website there isn't much more. You can certainly email questions, but I think the developer can do a lot better.
Don't expect the app to be able to reliably read handwritten notes. While the developer says the app can read a whiteboard, I haven't tried it and can't vouch for that function. Creaceed, the Belgian developer of Prizmo, reports it has submitted version 1.01 to the app store and is waiting for approval. Version 1.01 is supposed to fix a few bugs and improve "field handling" with business cards. That should fix the address field issue.
Everything else worked as advertised. To be carrying around such power in a pocket is truly amazing. I've already thought of a lot of uses. If I was in Germany, for example, I could scan a menu, translate the text to English, or have it read to me. I've already tried it on receipts, and grabbed some printed instructions I wanted to have in electronic form.
There are other programs that do OCR, other programs that do text to speech, and others still that will do translations. This is the first program that integrates these functions into a solid app. At $ 9.99 it has a premium price, but for those that need it, it's going to be indispensable.
Creaceed has a Mac version of Prizmo which is also excellent. Prizmo for iPhone works best on a 3GS or iPhone 4. It can work on a 2G or 3G iPhone if you use a Griffin Clarifi case. In the gallery, you can see Prizmo at work on a document during my testing.
