Probability Density Function of EMG Signals based on Hand Movements in Time and Frequency Domain

Abstract

This paper attempts to estimate the probability density function of hand movements by using EMG signals. Several hand grasps generated from different hand movements, we have analyzed Tip and Lateral. Four well known pdf functions for good fitness of test are Log Logistics (3P), Johnson, Dagum (4P) and Burr (4P), that have been tested. The probability density function has been carried out in time domain and FFT domain as well as in DWT domain. It was observed that there are different distributions for different hand movements, which describe the samples most accurately with the movements of hand with respect to two channels; channel 1 and channel 2. In this scenario, channels 1 is placed on upper limb and channel 2 placed on lower limbs as a reference channel. Although, Burr distribution and Log logistic distribution along with that Dagum has been a good fit for most of the data, it is shown in this paper that Non Negative distribution (Dagum (4P) and Burr (4P) distribution) is a better choice for estimating the Tip and Lateral hand movements. General TermsThis method is used for classification purpose. This is the another method of classification of data set in probability density function. This is used as a clinic/ engineering field as a designing of prosthetic arm and hand.