A method of recognizing finger-flexing motion using muscle action potential waveforms

Abstract

For the bedridden person's lives, not only the movement but also the communication have been restricted. Regarding our support of a daily life, the construction of the human interface system that substitutes the movement means and the means of communication becomes an important problem. Regarding the human interface system, the system device that can be easily used for many bedridden people is needed. Hence, we will propose the electromyogram interface system that have finger-flexing motion and can be easily moved even in the state of bedridden. The prototype model is actually constructed with an electromyograph of the telemeter type and a computer. The finger-flexing motion is identified by the discrimination analysis method in the data analysis. In this report, the prototype system is actually produced In the system, the finger-flexing motion can be recognizes more than 90% on average. It means that the reproduction of the means of communication can be used by this technique.