Orientation Prediction for VR and AR Devices Using Inertial Sensors Based on Kalman-Like Error Compensation

Abstract

We propose an orientation prediction algorithm based on Kalman-like error compensation for virtual reality (VR) and augmented reality (AR) devices using measurements of an inertial measurement unit (IMU), which includes a tri-axial gyroscope and a tri-axial accelerometer. First, the initial prediction of the orientation is estimated by assuming linear movement. Then, to improve the prediction accuracy, the accuracies of previous predictions are taken into account by computing the orientation difference between the current orientation and previous prediction. Finally, we define a weight matrix to determine the optimal adjustments for predictions corresponding to a given orientation, which is obtained by minimizing the estimation errors based on the minimum mean square error (MMSE) criterion using Kalman-like error compensation. Experimental results demonstrate that the proposed algorithm exhibits higher orientation prediction accuracy compared with conventional algorithms on several open datasets.

Keywords

• Orientation prediction
• inertial measurement units (IMUs)
• motion-to-photon (MTP) latency
• virtual reality (VR)
• augmented reality (AR)
• attitude and heading reference system (AHRS)