0

I'm interested in taking user stroke input (i.e. drawing with an iPad) and classifying it as either text or a drawing (or, I suppose, just non-text), in whatever capacity is reasonably feasible. I'm not expecting a pre-built library for this, I'm just having a hard time finding any papers or algorithmic resources about this.

I don't need to detect what the text is that they're drawing, just whether it's likely text or not.

aehlke
  • 15,225
  • 5
  • 36
  • 45
  • By text, do you mean a letter or a word? I saw a paper two years ago that described an algorithm to find regions in arbitrary images that contain text-like features. – Blender May 16 '13 at 16:55
  • I can't see how you can 'detect if it's likely text or not' without an attempt to actually recognize the text, so maybe you should see this: http://stackoverflow.com/questions/1697852/c-c-objective-c-text-recognition-library – piokuc May 16 '13 at 16:58
  • @piokuc I've seen evidence of it being done, I don't see why it would be impossible. It seems like a superset of actual text recognition. – aehlke May 16 '13 at 17:11
  • Also note that I have stroke input, not just a bitmap, so I have direction of strokes, etc. – aehlke May 16 '13 at 17:11
  • @aehlke I don't say it's impossible, just don't know how to do it :) Anyway, perhaps using one of the existing text recognition libraries is a way to go for you. – piokuc May 16 '13 at 17:40
  • [Lipi Toolkit](http://sourceforge.net/projects/lipitk/). might be useful – David Karlsson May 16 '13 at 20:12

1 Answers1

0

I would think you will need to generate probabilities for what text character the input is. If the highest probability text character is below some threshold, classify the stroke as drawing.

This is a possible useful paper: http://arxiv.org/pdf/1304.0421v1.pdf (if only for its reference list). Also the first hit on this google scholar search looks relevant: http://scholar.google.com/scholar?q=classification+stroke+input+text+or+drawing

Bull
  • 11,771
  • 9
  • 42
  • 53