Back to list

Sensor shield and fusion algorithm evaluation of DIY Attitude and Heading Reference Systems (AHRS)

Authors Berghold, P.
Year 2020
Thesis Type Audio Engineering project
Topic Spatial Audio
Keywords 3D sound, Ambisonics, binaural, Hardware, Human Computer Interaction (HCI), Motion Capture, signal processing
Abstract Head orientation estimation is an important task for plausible perception of virtual acoustics. To gain a reliable 3D acoustic scene via headphone playback, tracking of the listener's head movements is a necessary feature. Micro Electro Mechanical Systems (MEMS) oer a cheap and su cient possibility to track such movements, but lack in accuracy. This work evaluates state of the art sensor shields (BNO055, NXP 9DOF and MPU9250) and fusion algorithms for the de- sign of an attitude and heading reference system (AHRS) on the Arduino platform. The orientation estimation of the sensors is compared to an optical tracking system (Optitrack Flex 13). It is shown that all sensors can achieve su cient orientation estimation with the right fusion algorithm and proper calibration. Their perfor- mance diers in the robustness against magnetic distortions and gyroscope bias drift. Nevertheless the BNO sensor outperforms the others with its ease of use as it has an onboard sensor fusion.
Supervisors Rudrich, D.