Quadrant detector-based method for eye point alignment of augmented and virtual reality head mounted displays

Abstract

We demonstrate a quadrant detector-based method for aligning image sensors to the optical axis of augmented reality (AR) and virtual reality (VR) head mounted displays (HMDs). The sensor location that gives optimal image quality is known as the eyepoint. While a wide variety of methods have been explored for determining the eyepoint, the sensitivity and results of these methods depend on the optical design of the HMD. In our quadrant detector-based method, a test pattern consisting of dots is displayed on the HMD and imaged by a camera. The signal from the camera is divided into four quadrants and the camera is positioned at the eyepoint when the signal difference between the quadrants is minimized. This method has the advantage of giving the direction of the eyepoint from a single measurement without mapping the eyebox of the HMDs. The measurements were performed using an FDA developed setup for characterizing HMDs with the flexibility to incorporate a variety of light measuring devices and HMDs using commercially available components, which is described in detail. These results demonstrate a generalizable eyepoint alignment method, which allows for simpler automation of aligning the camera to the eyepoint.