Adaptive filtering algorithms and optical pose tracking for fully enclosed visualization spaces
Dirk Reiners, Louis Malcolm Hutson · 2011
Several critical limitations exist in the currently available tracking technologies for fully-enclosed virtual reality (VR) systems. While several six-degree-of-freedom tracking projects such as Hedgehog and HiBall have successfully demonstrated excellent accuracy, precision, and robustness within moderate budgets, these projects still include elements of hardware that can interfere with the user's visual experience and diminish the immersion in fully-enclosed VR systems. The objective of this project is to design a tracking solution for fully-enclosed VR displays that achieves comparable performance to available commercial solutions but without any artifacts that can obscure the user's view. JanusVF is a tracking solution involving a cooperation of both the hardware sensors and the software rendering system. A small, high-resolution camera is worn on the user's head, but faces backwards (180 degree rotation about vertical from the user's perspective). After acquisition of the initial state, the VR rendering software draws specific fiducial markers with known size and absolute position inside the VR scene. These virtual markers are only drawn behind the user and in view of the camera. These fiducials are tracked by ARToolkitPlus to produce observations of the camera's position in space. These observations are integrated by a single-constraint-at-a-time (SCAAT) adaptive filter algorithm to update the head pose. The filter provides sensor fusion, noise filtering, and state prediction. In order to improve the precision of the tracker during very slow movements and possible occlusions of the display surface, the filtering algorithm is extended to support inertial sensors including accelerometers and gyroscopes. The filter accounts for drifts and calibrates for changing biases of the sensor devices, updating the full state at 300 Hertz. Methods for optimizing the visualization, corresponding acquisition, and image processing are explored. Processes are developed to shade the virtual fiducials so that ambient light qualities are preserved. An alternate method is presented for reducing the total ambient light in a fully enclosed VR space by fading scene elements that are not visible to the user. Experiments analyzing accuracy, precision, and latency in a six-sided CAVE-like system show performance that is comparable to alternative commercial technologies.