Masanori Takemoto

The virtual image presented with Augmented Reality (AR) device is superimposed on the real world, and we can perceive it as if actually existing there. However, it is still unclear whether we could perceive a virtual image with the setting values for depth (distance and size). In this study, we examined the basic characteristics of depth perception for the AR images by the adjustment method (compared between an AR object and a real object) and the constant method (compared between an AR object and reference of distance). The results of experiments showed the overestimation of depth for a virtual image. In addition, the monocular observation would induce much overestimation than the binocular observation, while the sensitivity of depth improved with occlusion for another side of the eye where the virtual image was not presented. These results would help more appropriate presentation of the virtual image in the AR devices.