Abstract
In the paper, we proposed an anticipatory pattern recognition method for an electromyogram (EMG) prosthetic hand control. To detect humans' motor intentions as fast as possible, we focused on the transient state of the EMG signals. We sampled the EMG signals from a human forearm, and extracted a feature vector every 50 ms. At each time window, the proposed pattern discriminator determined whether the measured EMG signals were in the transient state or not In the case of the transient state, candidates of migratable prosthetic hand patterns were constrained by a prior knowledge regarding a transition probability among the hand patterns. Accordingly an appropriate hand pattern can be selected, otherwise an ongoing hand pattern was maintained. Online EMG pattern recognition experiments with the proposed anticipatory pattern recognition method demonstrated a significant improvement compared with the method without the anticipatory mechanism.
| Original language | English |
|---|---|
| Article number | 4811394 |
| Pages (from-to) | 897-902 |
| Number of pages | 6 |
| Journal | Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics |
| DOIs | |
| Publication status | Published - 2008 Dec 1 |
| Event | 2008 IEEE International Conference on Systems, Man and Cybernetics, SMC 2008 - Singapore, Singapore Duration: 2008 Oct 12 → 2008 Oct 15 |
Keywords
- Anticipatory constraint method
- Emg
- Prosthetic hand control
ASJC Scopus subject areas
- Electrical and Electronic Engineering
- Control and Systems Engineering
- Human-Computer Interaction