Purchase this article with an account.
Jeff D. Wurfel, Mónica Padilla, Norberto M. Grzywacz; Metric estimation of visual-deformation motions. Journal of Vision 2005;5(8):328. doi: 10.1167/5.8.328.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Optic flow has been shown to be used in navigation and self-orientation. Flow patterns caused by relative motions of small planar surface patches can be broken down into four elementary motion types: translation, rotation, expansion, and deformation (Koenderink & van Doorn, 1976). Many studies addressed the first three motion types. Results show that humans can finely discriminate between directions and speeds of translations (De Bruyn & Orban, 1988; McKee, 1981; Welch & Bowne, 1990; Bravo & Watamaniuk, 1995), angular velocities of rotations (Barraza & Grzywacz, 2002 & 2003), and rates of expansion (Wurfel, Barraza, & Grzywacz, 2003). In contrast, deformation and its rate have been studied comparatively less than the other motion types. Using a 2AFC test, we show that subjects can also finely discriminate between rates of deformation. Moreover, we show that subjects can discriminate accurately between orientations of the axis of deformation. Another similarity between deformation and the other elementary optic-flow types is in the quantity of information needed for fine discrimination. A certain amount of motion information is required to make angular-velocity and rate-of-expansion judgments in rotation and expansion respectively. The rate of deformation is also subject to similar motion-information constraints. In other words, deformation stimuli with few moving elements will typically be judged through local speeds, whereas deformations with many moving elements will be judged by the global rate of deformation. Finally, we show that the rate of deformation can be discriminated with short stimulus durations (155ms), like the other elementary motion types. In conclusion, although deformation is a less intuitive type of optic flow than translation, expansion, and rotation, it seems to be discriminated using a similar strategy. This supports the hypothesis that the brain decomposes optic flows into the components derived by Koenderink and van Doorn.
This PDF is available to Subscribers Only