Is there a pen scanner that does OCR on handwritten text?

[kranky]

I'm trying to find a pen scanner that does OCR on hand-printed block letters (not just on printed text). It would be something like this one, except the QuickLink does not do OCR on hand-printed text.

The idea is to scan/OCR neatly printed block letters and have the pen enter them into an application just like it came from the keyboard. The QuickLink does all that except it only works on preprinted text, not hand lettering.

Suggestions are greatly appreciated.

 

[Pennstate]

The irony is that some pen-based chinese-character recognition software also does english recognition. My dad has one and it works in MS word.

 

[corkyg]

That is because hand lettering is a variable - and OCR depends on consistency of form for accuracy in recognition. Because of all the editing involved, it is usually faster to just type copy the text. Even the old hand scanners were not very good for OCR work.

A different, and perhaps easier approach to consider would by voice translationj into text - i.e., read the text alound into a Via Voice type program.

 

[kranky]

I understand the need for consistent lettering for OCR to work. The type of printing I want to use this on will successfully OCR from a regular scan/OCR program like TextBridge (at least as much as you can expect from OCR), so it's tantalizing to hope that there's a way to do it with a pen unit.

It seems like the pen-based units maybe don't have the horsepower to do a quality OCR job. The application doesn't lend itself to doing a bunch of text scanning, then a batch OCR. I need to scan a few characters, have the device OCR it and send it along to the Windows app as if it came from the keyboard. If I have to process the scanned text through a big OCR program in a big batch, it's not worth it.

The reason I don't want to retype (even though the text strings are pretty short - like 3-8 characters) is that the text strings are essentially random letters, and it is very difficult to type them accurately as compared to regular sentences.