An Exploratory Study of Single Channel Surface Electromyography for Hand Gesture Classification

Abstract

Accurate hand gesture recognition using surface
electromyography (sEMG) typically relies on multichannel sensor
arrays and computationally intensive models, limiting practical
deployment in low power and embedded systems. This study
investigates the feasibility of classifying ten hand gestures using
a single sEMG channel combined with lightweight machine
learning models. Raw sEMG signals were transformed into
a comprehensive feature based representation, including time
domain, frequency domain, higher order crossing, and rela-
tive intensity features. Feature redundancy was reduced using
Pearson correlation filtering and removing highly correlated
features, while dimensionality reduction techniques (LDA and
PCA) were applied selectively. Three classifiers—feed forward
neural network (NN), k-nearest neighbors (KNN), and sup-
port vector machine (SVM) was systematically evaluated across
four experiments. Results demonstrate that a carefully designed
feature based pipeline, particularly when combining time and
frequency features with Pearson filtering and a compact NN,
can achieve up to 90% accuracy, even with limited temporal
and spatial information. These findings highlight the potential
for single channel sEMG systems in cost effective, low power
gesture recognition applications.