Controlling Mobile Robot Using IMU and EMG Sensor-Based Gesture Recognition

Abstract

In recent years of the field of robotics, various methods are developed for intimate relationship between people and robot. These methods include speech recognition as well as gesture-based interaction. This paper proposes a system that controls a mobile robot by recognizing the user gesture. The proposed system applies hidden Markov model (HMM) using inertial measurement unit and a single electromyography (EMG) sensor. HMM is known as the rich formula system, which can be applied to many applications including the gesture recognition. HMM is simply not only a gesture, but also can be applied to human robot interaction (HRI) very effectively. Thus, this paper uses the HMM to generate a familiar interaction between user and mobile robot. The body gesture is recognized using IMU sensor, which is easy to use for anyone. The mobile robot is then controlled according to the continuous incoming EMG command signal through Bluetooth, which is determined by the user gesture. Lastly, the HMM is applied as the method for determining the accuracy of user posture. In this way, more natural and intelligent gesture-based control system is implemented.