Gesture Recognition from Sequence of Difference Images
Various kinds of position detecting devices have been designed so that users
can immerse themselves into virtual environments.
So far, semantical operations are realized via command selection on the menu.
In this paper, the gesture interface is proposed which provides the application
program with gestical information such as the type, the relative speed and the
extent of user's gesture.
Experimental results demonstrate that the type and the direction of user's
gesture are recognizable even under lower frame rates, but higher frame rates
are necessary in order to obtain the relative speed and the extent of the user's gesture
with more accuracy.
Last update : March 8, 1996
Toshiyuki Kirishima(tosiyu-k@is.aist-nara.ac.jp)
