Recognition of Sign Language Numbers via Electromyography Signals


Ketenci S. , KAYIKÇIOĞLU T. , GANGAL A.

23nd Signal Processing and Communications Applications Conference (SIU), Malatya, Turkey, 16 - 19 May 2015, pp.2593-2596 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • Doi Number: 10.1109/siu.2015.7130416
  • City: Malatya
  • Country: Turkey
  • Page Numbers: pp.2593-2596

Abstract

Muscle signal is widely utilized in recognition of hand gesture, prosthetics and rehabilitation. Some study is available to recognize the hand gesture via EMG for some sign languages. Sign language performed generally with hand movement is developed for deaf. According to our research, any study is not available for numbers of between 0 and 9 in Turkish sign language. In this paper, surface electrodes put on fore arm were used to record EMG signals in order to determine these numbers. Features were extracted using root mean square, variance, waveform length, Fourier transform coefficients and proposed standard deviation of crosscorrelation coefficients after preprocessing. It caused that performance of linear discriminant analysis increased highly.