Study of stereoscopic shape perception for virtual design

Virtual environments are increasingly used to replace real prototypes, but are we sure that shape perception is the same in both environments, i.e. can we affirm that the shapes that are designed and validated in virtual environments will be perceived the same once a physical prototype has been built? This study traces a series of perception tests in which two virtual shapes (a sphere and a shape without symmetry or obvious monoscopic cues) will be compared to their respective actual shape. Projections change throughout the tests to highlight the influential parameters. We compare influence of head tracking, adjusting interpupillar distance, position of the virtual object in relation to the screen and the orientation of the modified shape. We show that if all these parameters seem to have an influence, head tracking appears extremely important for a good perception.