December 2022
Volume 22, Issue 14
Open Access
Vision Sciences Society Annual Meeting Abstract  |   December 2022
Estimating Object Depth from Optic Flow: Comparing Optical Expansion and Motion Parallax
Author Affiliations
  • Siyun Liu
    University of Minnesota
  • Gordon Legge
Journal of Vision December 2022, Vol.22, 4050. doi:https://doi.org/10.1167/jov.22.14.4050
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Siyun Liu, Gordon Legge; Estimating Object Depth from Optic Flow: Comparing Optical Expansion and Motion Parallax. Journal of Vision 2022;22(14):4050. https://doi.org/10.1167/jov.22.14.4050.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose. An observer’s motion can produce depth cues that are useful in judging object locations during spatial mobility. Based on the optic flow pattern, motion-related depth cues can be categorized as optical expansion when the observer approaches the object, and motion parallax when the observer’s motion is orthogonal. We asked whether optical expansion or motion parallax yielded more accurate estimation of relative object depth, and whether static views before and after a motion sequence were as effective as continuous motion in providing depth information. Method. Stationary observers were tested in 240 trials. They viewed videos of a virtual room containing a hanging hexagonal sign as reference, and a cube on the floor that ranged from six ft in front to six ft behind the sign. Subjects used an on-screen slider to estimate the depth of the cube relative to the sign after 2-second videos simulating optical expansion or motion parallax with the same camera translation distance, or a static image pair showing the first and last frames of the motion videos for 0.5 seconds each. Ten subjects were tested remotely using the PsychoPy software during quarantine, and later ten in-person in our lab. Results. Motion parallax produced significantly smaller depth estimation error (M = 1.07 ft for remote viewing, M = 1.33 ft in-lab) than Optic expansion (M = 2.33 remote, M = 2.46 in-lab). Static image pairs were not significantly different from continuous motion in the remote test but produced higher estimation error than continuous motion in the in-person tests. Conclusion: Motion parallax yielded more accurate estimates of object depth than optical expansion. Static image pairs yielded less accurate estimates than continuous motion for in-person tests.

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×