The rapid progress of integrating powerful processing units into embedded devices has enabled research groups and the industry to develop interaction devices that provide much more intuitive usage, thus increasing the popularity of virtual reality technologies in recent years. Evaluating the results, we may find promising concepts, which is where this paper intends to add its contribution.
The goal of this work is to take the users' needs for intuitiveness and flexibility with the implementation of a compact device into consideration, the purpose of which is to provide real-time orientation and position estimation in three dimensions. As most significant factors we name reliability and power-saving operation mode.
The current work has its roots in the in 2017 SSC paper presented at the Budapest University of Technology and Economics, which has provided an analysis of the prototype's bottlenecks, thus grounding the development of the new hardware and software architecture. The new device was also presented in an SSC paper this year, the current document extends the scope of it mainly regarding implementation aspects.
The interface device presented in the current paper consists of a nine degrees of freedom MEMS inertial measurement unit, the measurements of which are transmitted wirelessly, while the data processing is done by microcontrollers of the PIC family. To make the integration seamless for the user, the HID Device Class of the USB standard was applied.
Orientation-estimation is based on the Madgwick-algorithm, a problem-specific quaternion-based solution, while for the position estimation Hidden Markov Models are used. The latter is the more challenging task, due to the noise processes within the accelerometer - to handle this problem with a reasonable complexity for the embedded environment, some simplifications were used.
Structurally, the paper is organized as follows: after the introduction of the existing solution the theoretical background is shortly summarized. Previous and current work is separated, and the implementation/measurement details are elaborated.