Machine Learning-Based Classification of Neck Movements Using sEMG and STM32 Microcontroller

---

Abstract

In this paper, we propose a machine learning system for the classification of neck movements from sEMG signals. Ten subjects were measured with four Bio Amp Patchy sensors, which were attached onto the Sternocleidomastoid, upper trapeziusm and scalene muscles. A STM32F411RE microcontroller was used for timing and accurate acquisition of the EMG signals. The methodology consisted of five stages: data collection, signal preprocessing, feature extractions, classification, and performance assessment. Time domain features were also extracted and classification using a Random Forest algorithm. Finally, high accuracy was achieved and the system demonstrated great potential in posture monitoring, rehabilitation, and hands-free human–machine interaction

---

Keywords

Electromyography (EMG), classification of neck movement, microcontroller STM32F411RE, Bio Amp Patchy Sensor, Random Forest Algorithm, machine learning, Python, Sternocleidomastoid (SCM), Upper trapezius, and Scalene muscles.