Based Classification of Tip and Lateral hand movements based on Time Frequency domain

Abstract

This paper portrays a method for improvement of the accuracy of hand movements based on EMG signals in time and frequency domain and to reduce the misclassification of Tip and Lateral hand
movements has been pro- posed. ANN is trained by the trained dataset and testing is done by test data. Principal component analysis method with ANN has implemented to reduce dimensionality of
the large dataset and after that train the artificial neural network. To reduce the misclassification complexity of the proposed algorithm, classification methods may be employed. It can include
classification methods as Support Vector Machine. Support vector machine classifier is also implemented for improve the accuracy of the two complex hand movements such as Tip and Lateral
movements based on electromyography signals. The accuracy that has improved in this present work is near about 55%. This work is beneficial in future for those people who lost their hand correctly
control ability, and for those whose body part is replaces with prosthetic devices such as forearm, hand. By using this technique we improved the accuracy of prostheses device. It will help the
clinicians/Biomedical to diagnose the patients effectively and also for the engineers to design the prosthetic limp easily and correctly