Compression of electromygraphic signals using adaptive predictive multi-rate codec / Compressão de sinais eletromiográficos usando codificador multitaxa preditivo adaptativo

AUTOR(ES)
DATA DE PUBLICAÇÃO

2009

RESUMO

The development of highly efficient speech codecs has raised the interest in their use in other types of signal, such as electrophysiological ones. This study evaluates the AMRACELP codec, developed for third generation mobile telephony, and proposes changes to improve its performance in compression of electromyographical signals, considering the objective quality of reconstructed signal. Such changes were used in several new modes of operation of the codec, which was evaluated using EMG signals from isometric contractions. There were obtained compression rates from 81.4% to 90.3%, and a minimum mean square error of 1.81%. It was found that the objectively measured quality of the signal increases with the exclusion of the elements of subjective quality enhancement and the adaptive dictionary and with the inclusion of a second fixed dictionary. The results are consistent, in terms of compression rate and error, with previous works on compression of EMG under similar conditions, using wavelets or the AMR codec. Some of the new modes were more efficient, under the mentioned criteria, than the previous work using AMR.

ASSUNTO(S)

codificação eletromiografia compressão processamento digital de sinais amr engenharia eletrica emg acelp

ACESSO AO ARTIGO

Documentos Relacionados