Cepstral and Non Uniform Filter Bank Features for Processing EMG Signal in Person Identification

The identification of a person using biometric data may be of interest. A direct connection between muscle, central nervous system and brain is unique to an individual. Here we have demonstrated the significance of electromyogram (EMG) signal for person identification using cepstral and non uniform filter bank features. Our approach consists of a robust feature extraction scheme which is based on cepstral analysis and non uniform filter bank with a specified configuration, combined with vector quantization modelling. Various experiments have been conducted to determine the person identification performance in our proposed scheme. Preliminary results indicate that with well-chosen feature extraction, an identification rate of up to 93.8766% is achievable for a database consisting of forty nine individuals, with collected 3 sessions EMG data in a gap of one day duration.