Tracking and classification of head movement by inertial measurement unit data

Abstract

Subject of Research. We present the algorithm for tracking and classification of the operator's head movement according to the data obtained from an IMU (inertial measuring unit) installed in the helmet of virtual reality. We performed experimental testing of the developed algorithm and proposed the way of its application for control of mobile robot with six degrees of freedom. Method. The problem is solved by complex filtering of input signals from the gyroscope with the Kalman filter and the dead-zone filter. The definition of the operator's head movement pattern is based on determination coefficient calculation and its comparison with the threshold level. Main Results. We have developed an algorithm for unambiguous evaluation of the operator's head movement that can be interpreted by the control system as a control signal. Practical Relevance. Presented algorithms can be used in a wide range of systems, for example, for controlling mobile telepresence robots that have special suspension for camera guidance control and are controlled with the use of virtual reality helmet. The developed algorithm will be applied also in the implementation of mobile robotic platform, equipped with computer vision systems, navigation and augmented reality, that is planned to be created in the framework of practice-oriented research and development work at ITMO University.