How to develop a OCR application that can scan documents online?

How to develop a OCR application that can scan documents online?

Do you mean something like this

Displeaser

yes i mean something like this
how should i go about working on something like this?

Afraid I've never used it so wouldnt know

The SDK help will go through the technical explanations of each of the APIs functions and there should be some example code either in one of the downloads or on the site. If you cant find any then drop their help a mail and they should be able to give you more assistance.

Displeaser

so u mean the sdk ..i can just use the php or watever language to call on the packages?

I had a look into and it seems its only for turning paper documents or tiffs into electronic format. Is that what you are planning on doing?

For programming and usage it seems to only provide a C++ and an activeX wrapper.

Displeaser

hmm i c..so is there anyway i can use php to call it..anyone have any idea?

what exactly are you trying to do?

If its parse html/xml pages then Perl or another similar language will do that easily for you. If your trying to parse the text from gifs/jps etc then you can use the above sdk (after you convert them to tiffs).

Give us a bit more info to work with.

Displeaser

You can use ocrsdk.com for ocr development of your application. See documentation section for details.

There are few parts

* Web application that allows users to upload image (create a cookie with unique ID)
* Images are stored in a "processing" folder (possibly marked by a unique ID)
* A separate service application on the server that takes batches of Images from the "processing" folder.
* Service application OCR's an image file and then places the data into a database (with the ID). Original File is deleted.

* Web application polls database with id from Cookie.
* When data becomes available, result is rendered as HTML. Cookie is deleted, data result record is deleted.

In order to create the service application you will have two options:

* Use an existing OCR library.
* Create your own OCR engine.

I hope that helps.