Implementación de un algoritmo para el reconocimiento de gestos del brazo humano en tiempo real utilizando la Imu del dispositivo Myo Armband

Abstract

In this work, an algorithm for the recognition of 8 movements of the arm is implemented using the IMU of the Myo Armband device placed on the forearm. The movements to recognize are: left, right, up, down, clockwise, counterclockwise, backward and forward, depending on the initial position of the forearm, which can be horizontally or vertically. Gesture recognition is done through three stages: the first one deals with the acquisition of the signals where the pertinent configurations are made so that there is good wireless communication, the appropriate communication ports are opened and they are collected within vectors. The second stage deals with the processing of the signals where the collected data are processed independently by making the appropriate transformations with the mathematical formulas already established for this process. Subsequently, the specific characteristics in each movement are searched to be classified depending on the three-dimensional plane of movement that is compared with respect to a fixed reference system. Finally, the results are published on the computer screen in real time, reflecting an image of the movement carried out and its respective na