Looking for guidance on multi-part graphical gesture recognition

Question

I'm trying to research existing works in the area of recognizing complex graphical gestures, but struggling to find good search terms or clear documents in the field.

For example, I might want to recognize an arbitrary shape drawn by a mouse as a path, or a path with an arrow head, or an enclosed region, or some combination of various shapes. One of the problems I'm having is that the shapes are not necessarily going to take a consistent symbolic form, so I'm looking for alternate ways to represent the data.

This article has been helpful so far, but does not discuss any concrete algorithms. http://www.springerlink.com/content/l4235612gv479262/

Anyway, I'm not necessarily looking for solutions here (though I'd be interested if there were some), but if people can help direct me towards some pertinent papers or the right kinds of search terms, I would appreciate it!

Answer 1

If you take the sequence in which the gesture is made/drawn into account then you might want to make use of sequence alignment techniques. Asking Google gives quite a few matches which you might want to explore.