Optical Character Recognition

[ca8msm]

Hi,

I was wondering if anyone has had any experience with Optical Character Recognition (OCR) in .NET? We are looking at the possibility of providing an application that can take a scanned TIFF image and gather the text from that image. We would then do our own processing with the text that was returned.

I had a play about with the Microsoft Office Document Imaging application as there is a COM component that could have been used but it returned some pretty poor results as it didn't correctly interpret a lot of the text that I though it would.

If anyone does have any experience with it, are there any recommendations you can make (3rd party components are fine and cost isn't that much of an issue) it would be much appreciated.

Thanks
Mark

----------------------------------------------------------------------

Need help finding an answer?

Try the search facility (

http://www.tek-tips.com/search.cfm)

or read FAQ222-2244 on how to get better results.

 

[ThatRickGuy]

When in doubt, check

http://SourceForge.net

Search for OCR listed 3 pro tools and numerous free/GPL solutions.

-Rick

----------------------

http://www.ringdev.com

I believe in killer coding ninja monkeys.

 

[SqueakinSweep]

I'd also be interested in hearing from anyone who has actually used 3rd party controls, and successfully integrated them, for a future Document Management System I'm going to be landed with.


Sweep
...if it works dont mess with it

 

[ca8msm]

Thanks Rick.

Some of the sponsored links are for scanning services rather than software but I'm trying out the "Readiris Pro 9 OCR" application now.

In case anyone else reads this thread (and for future reference for Sweep) then the best results I've had so far have been from a product named SimpleOCR (

http://www.simpleocr.com/)

but I'm still not happy with any of the results I've seen so far.

I'll keep this thread updated if anything turns up.

mark

----------------------------------------------------------------------

Need help finding an answer?

Try the search facility (

http://www.tek-tips.com/search.cfm)

or read FAQ222-2244 on how to get better results.

 

[ca8msm]

Right here's another update:

I've downloaded and tested the "Readiris Pro 9 OCR" (it's actually version 10 for the windows version) from

http://www.irislink.com

and it is way ahead of any of the other products.

I tested it on a document that contained Arial text and it converted every single letter!

On a document that had some hand-written text it didn't fair as well (although that is to be expected) but it still beat other products hands-down.

I would definately recommend this product as the best I've seen so far.

Thanks
mark

----------------------------------------------------------------------

Need help finding an answer?

Try the search facility (

http://www.tek-tips.com/search.cfm)

or read FAQ222-2244 on how to get better results.

 

[ThatRickGuy]

Good to know!

-Rick

----------------------

http://www.ringdev.com

I believe in killer coding ninja monkeys.

 

[chrissie1]

and how does it work with .net???

as com??

Why didn't anybody respond to my question when I asked this????




Christiaan Baes
Belgium

If you want to get an answer read this FAQ faq796-2540
There's no such thing as a winnable war - Sting

 

[ThatRickGuy]

Chrissie, you can still use COM objects in .Net. They might make you want to swear, but you can still use them. And as for the replies, I just think ca8msm is cooler

-Rick



----------------------

http://www.ringdev.com

I believe in killer coding ninja monkeys.

 

[chrissie1]

I just think ca8msm is cooler

Not really, about 36,5°C.

I know you can use com but I like .net better. I just want somebody to test it before I buy it.

Christiaan Baes
Belgium

If you want to get an answer read this FAQ faq796-2540
There's no such thing as a winnable war - Sting

 

[ca8msm]

I'll see if I can give it a quick test when I get to work tomorrow.

Oh, and you're right - I am cooler than chrissie

----------------------------------------------------------------------

Need help finding an answer?

Try the search facility (

http://www.tek-tips.com/search.cfm)

or read FAQ222-2244 on how to get better results.

 

[ca8msm]

The product itself doesn't allow you direct access but on their site there is a ".net integratable toolkit" that supports direct integration in any development of "VB.NET, C#, C#.NET, J#, J#.NET" so it looks as though it should be quite good.

----------------------------------------------------------------------

Need help finding an answer?

Try the search facility (

http://www.tek-tips.com/search.cfm)

or read FAQ222-2244 on how to get better results.

 

[chrissie1]

BTW Iris is a belgian firm.

Christiaan Baes
Belgium

If you want to get an answer read this FAQ faq796-2540
There's no such thing as a winnable war - Sting

 

[ca8msm]

That'll be why it's good then is it Chrissie?!

----------------------------------------------------------------------

Need help finding an answer?

Try the search facility (

http://www.tek-tips.com/search.cfm)

or read FAQ222-2244 on how to get better results.