Characterization and correction of spatial misalignment in head-mounted displays

Date of Award


Degree Name

M.S. in Electro-Optics


Department of Electro-Optics and Photonics


Advisor: Andy C. Chong


A toolset was developed for characterizing and correcting spatial misalignment in head-mounted displays. A hardware system consisting of two cameras and various rotation and translation stages was used to emulate the ocular position of most human observers. A checkerboard pattern was displayed on the HMDs and matched to a reference pattern through an image registration process. The HMD image registration process is carried out after the effects of camera distortion and keystone effect are removed. The registration process is repeatable with a standard deviation of less than one HMD pixel.The relative misalignment between left and right eyes was fairly small in the center of the displays, and increased near the edges and corners. Small rotations simulating an imperfectly aligned HMD had little effect on the misalignment present. The introduction of vergence angles did have a large effect on misalignment. Several methods were used to correct misalignment, including different corrections for the left and right eyes, and the use of a composite correction incorporating different correction maps in different local regions of the display. Both of these methods showed improve uniformity and rectilinearity in test images displayed on the HMD. The composite correction map did show noticeable global variations. The luminance of an HMD was also characterized, showing higher luminance in the center of the display than in the corners by a factor of three.


Virtual reality headsets, Image registration, Vergence (Binocular vision), Optics, head-mounted displays, HMD, misalignment, optical aberrations, image registration, distortion, luminance, virtual reality, augmented reality, biocular, binocular, human factors, vision, vergence, eyestrain, visual fatigue, symbology, aviator

Rights Statement

Copyright © 2017, author