Hand Movement Classification Using Burg Reflection Coefficients

Abstract

Classification of electromyographic signals has a wide range of applications, from clinical diagnosis of different muscular diseases to biomedical engineering, where their use as input control of prosthetic devices has become a hot topic of research. Challenge of classifying this signals relies on the accuracy of the proposed algorithm and the possibility of its implementation on hardware. This paper consider the problem of electromyography signal classification, solved with the proposed signal processing and feature extraction stages, with focus lying on the signal model and time domain characteristics for better classification accuracy. The proposal considers a simple preprocessing technique that produces signals suitable for feature extraction, and the Burg reflection coefficients to form learning and classification patterns. These coefficients yield a competitive classification rate compared to used time domain features. Sometimes, the feature extraction from electromyographic signals showed that procedure can omit less useful traits for machine learning models. Using feature selection algorithms provides a higher classification performance with as fewer traits as possible. Algorithms achieved a high classification rate up to 100% with low pattern dimensionality, with other kinds of uncorrelated attributes for hand movement identification.

Evaluation of Feature Extraction and Classification for Lower Limb Motion Based on sEMG Signal

“…The feature extraction methods of the sEMG signal mainly include time domain, frequency domain, and time-frequency domain. Among them, time-domain analysis is the most commonly used method such as integrated sEMG (IsEMG), mean absolute value (MAV), simple squared integration (SSI), root mean squared (RMS), wavelength (WL), zero-crossing (ZC), and Willison amplitude (WAMP) [ 13 ]. FEIYUN XIAO et al used root mean square, waveform length, the absolute standard deviation of difference, integrated sEMG signal (IsEMG), and sEMG low-pass filtered (50 Hz) signal (LPFEMG) features to quickly and accurately estimate joint motion [ 14 ].…”

sEMG Signals Classification using CNN Features Extraction as a Reliable Method

“…Finally, they reduced the size of the problem through Principal Component Analysis (PCA). Another work that analyzed this database was Ramírez-Martínez et al (2019). The authors innovated by adding features in the time domain and a methodology called Burg Reflection Coefficients to obtain more information from the signals.…”

Heterogeneous Hand Guise Classification Based on Surface Electromyographic Signals Using Multichannel Convolutional Neural Network

“…Later that year, Iqbal, Fattah and Zahin used SVD and PCA for feature extraction and selection, and then 𝑘-NN for classifying the figure position classes [9]. In 2019, Ramírez-Martínez et al used Burg reflection coefficients to build an optimal feature set, and carried out a series of classification tasks using different classifiers and preprocessing techniques [10]. In the same year, Nishad et al designed a cost-effective sEMG classification method applying Tunable-Q Wavelet Transform (TQWT) based filter-bank for signal decomposition, Kraskov Entropy (KRE) for feature extraction, and finally, 𝑘-NN for carrying out the classification of various hand movements [11].…”