Habiba Aziz 26 - Academia.edu (original) (raw)

Uploads

Papers by Habiba Aziz 26

Research paper thumbnail of Hand Motion Recognition from EMG using Artificial Neural Network

Hand motion recognition has become an active research due to its numerous applications such as it... more Hand motion recognition has become an active research due to its numerous applications such as its use in human-computer interface. The motivation for this work is to help the disabled people by improving their quality of life. This paper aims to recognize and replicate four hand gestures fist, spread, wave in, and wave out on 3D printed prosthetic hand. Electromyography (EMG) signals are recorded for these gestures using Myo armband consisting of eight electrodes from which five statistical parameters of EMG signals are extracted and employed for classification. These parameters for each electrode accumulate to form feature vector inserted to Artificial Neural Network (ANN) which classifies it into its target classes (gestures). The performance of ANN classifier is assessed over Scaled Conjugate Gradient (SCG) in comparison of different algorithms. Our simulation results are also supported with experimental results run over 3D printed prosthetic hand.

Research paper thumbnail of Hand Motion Recognition from EMG using Artificial Neural Network

Hand motion recognition has become an active research due to its numerous applications such as it... more Hand motion recognition has become an active research due to its numerous applications such as its use in human-computer interface. The motivation for this work is to help the disabled people by improving their quality of life. This paper aims to recognize and replicate four hand gestures fist, spread, wave in, and wave out on 3D printed prosthetic hand. Electromyography (EMG) signals are recorded for these gestures using Myo armband consisting of eight electrodes from which five statistical parameters of EMG signals are extracted and employed for classification. These parameters for each electrode accumulate to form feature vector inserted to Artificial Neural Network (ANN) which classifies it into its target classes (gestures). The performance of ANN classifier is assessed over Scaled Conjugate Gradient (SCG) in comparison of different algorithms. Our simulation results are also supported with experimental results run over 3D printed prosthetic hand.

Log In