EMG-driven movement decoding and control of an humanoid robot

Implementations of full robotics controls derived from biological signals, like EMG, are still not fully accomplished. Therefore, the aim of this thesis is to create a reliable system, for real-time applications, able to decode and then replicate on a humanoid robots the same motor task of a subject, using as an input only EMG signals. This study is divided in two parts: target-oriented classification and kinematic reconstruction analysis