Pen-Based Gesture Recognition in Multimodal Human-Computer Interaction
Yan Wang · 2001
This paper proposes a fast, one stroke pen gesture recognition approach to the studying of multimodal human computer interaction theory and building method. In the approach, a pen gesture is characterized by a sequence of dominant points along the gesture trajectory and a sequence of writing directions between consecutive dominant points. The recognition result can be obtained by matching the feature code of the input gesture with the various possible feature codes of each standard gesture. The directional feature is used for gesture pre-classification and the positional information is used for fine classification. Experimental results show that this approach is fast and can get a high recognition rate.