If I understand correctly, you want to scan a paper document,
and recognize properties (Such as text and images).
Then load them into a computer.
Theoretically it is possible, however the technology is very abstruse.
You would have to use the internal camera on the device to take a picture of the document.
Then map it with landmarks (Such as corners and paragraphs).
The factors would be endless (Light, Resolution, Angle, etc.)
I recently made a desktop application with the webcam library.
It took a picture every 10th second, and if the picture was any diffrent than the last one,
an alarm would go off (Good for burglers
)
My point is, you might have to rethink the idea.
Of course, I may not have understod correctly :sign0161:
Article about iPhone app that scans checks using a camera:
Bank Will Allow Customers to Deposit Checks by iPhone