UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Calibrating head-coupled virtual reality systems Stevenson , Alexander

Abstract

Head-tracking virtual environments are difficult to implement because of the need to calibrate such systems accurately, as well as the difficulty in computing the correct off-axis image for a given eye location. The situation is further complicated by the use of multiple screens, the need to change the calibration for different users, and the desire to write portable software which can be reused on different hardware with varying screen configurations. This, thesis presents a solution to these problems, allowing greatly simplified development of head-tracking software. By making use of-the head-tracking sensors built into the environment, we can quickly and accurately calibrate not only userspecific measurements, such as eye-positions, but also system measurements, such as the size and locations of display screens. A method of doing this calibration is developed, as well as a software library which will read a system configuration and integrate with OpenGL to compute correct off-axis projections for a user's viewing position. The calibration makes use of a novel "sighting" technique which has the great advantage of accurately finding the true rotational centre of a user's eyes. To complement this, the software library includes functions which predict the optical centre of a user's eye based on a given gaze point. As a demonstration of both the calibration method and the utility library, a hardware rendering application is discussed. This application performs the real-time rendering of view-dependent LaFortune reflectance functions in graphics hardware. As with all view-dependent lighting methods, both the viewing angle and position of the light are taken into account while rendering. Head-coupling allows the system to use the user's true viewing direction in the lighting computation, and the position of the virtual light is controlled by a 3D sensor in the user's hand. The method in which the view-dependent lighting model is implemented in hardware is explained, as well as possible improvements. Throughout, the Polhemus FASTRAK is used as the tracking system, though all the results are easily applicable to any six degree-of-freedom tracking system.

Item Media

Item Citations and Data

Rights

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.