We examined errors in the localization of nearby virtual objects presented
via see-through helmet-mounted displays as a function of viewing conditions
and scene content in four experiments using a total of 38 participants. Mo
nocular, biocular, and stereoscopic presentation of the virtual objects, ac
commodation (required focus), participants' age, and the position of physic
al surfaces were examined. Nearby physical surfaces were found to introduce
localization errors that differ depending on the other experimental factor
s. These errors apparently arise from the occlusion of the physical backgro
und by the optically superimposed virtual objects, but they are modified by
participants' accommodative competence and specific viewing conditions. Th
e apparent physical size and transparency of the virtual objects and physic
al surfaces, respectively, are influenced by their relative position when s
uperimposed. The design implications of the findings are discussed in a con
cluding section. Head-mounted displays of virtual objects are currently bei
ng evaluated as aids for mechanical assembly and equipment maintenance. Oth
er applications include telesurgery, surgical planning, telerobotics, and v
isualization aids for robotic programming.