Abstract
This study proposed robust EMG measurement technology, and examined the motion identification based on the AIEMG feature from a subject's forearm surface EMG. Several subject's motion identification experiment was conducted, and it was confined that the form of fingers could be identified with the probability of 90% or more.