Japanese sign-language recognition based on gesture primitives using acceleration sensors and datagloves
Hideyuki Sawada, Takuto Notsu, Shuji Hashimoto · 1998
This paper proposes a Japanese sign-language recognition system using acceleration sensors, position sensors and datagloves, to understand human dynamic motions and finger geometry. The sensor integration method realized a robust gesture recognition comparing with a single sensor method. The sign-language recognition is done by referring to a Japanese sign-language database in which words are written as sequences of the gesture primitives. Two recognition algorithms which are the automata algorithm and the HMM are introduced and tested in the practical experiments.