Using Windows Tablet PC Input to Implement Handwriting Recognition
I want to write an application (originally Windows) that includes handwriting for text recognition. I want to use the built-in Windows Tablet PC INput. My question is how to capture strokes as an image, "send them to the OCR engine used in Tablet Input and return the recognized text?"
Or are there any good open source post libraries that can be used directly?
The main development language is Qt.
source to share
I am not aware of any open source or free handwriting recognition software libraries, so I wrote an adapter. My goal was my Linux tablet PC, but part of my solution can also be used directly on Windows, although you will need to adapt it to your needs.
You will need to review the licenses for the components I have used and verify their own usage.
Source available here: Ink2Text Project
Part of this solution is a server that uses XP handwriting recognition libraries to interpret the strokes that make up handwriting. As an aside, it doesn't use OCR - it uses associated stroke flow plots.
Another additional project provides a widget for client handwriting: Stylus / Handwriting Input Panel . This is written in Java and this is GPL3. It takes the handwriting and sends it to the server. If you don't want to use it as is, it is solely relevant for viewing the data format for ink, although it is simple enough and you can probably conclude that only with the source code of Ink2Text.
An early solution was using S / HIP with my MS Ink Server , which accepts logins through regular network connections. This can also be useful depending on your architecture, but it requires a working copy of Windows.
This system provides very good recognition of typed and cursive handwriting.
I will only answer questions about this in the related SourceForge forums so others can benefit from the answers, please do not ask here.
Cheers, Bret
source to share
I want to be wrong, but unfortunately there is no open source standalone handwriting recognition system available even close to MS or Apple Ink.
On Windows, you can play with ink recognition ( About handwriting recognition , Advanced Recognition Example ). The C ++ interface is available, but not as well documented as the .net implementation. So, you need to put in more effort and do a lot of research to achieve what you want.
For other systems (including Windows), there is a way to use Tesseract-OCR with your application. See Tesseract base api . For better recognition quality, you can train tesseract and use your own prepared data.
If you don't want to waste time doing the R&D tasks above, you can use paid solutions like: MyScript SDK , WritePad SDK , etc.
source to share