Stone LS, Perrone JA.
Invest Ophthalmol Vis Sci. 1993 Mar 15; 34: 1229.
NASA Ames Res. Ctr., Moffett Field, CA, USA.
Purpose. We have already shown that humans can make reasonably accurate heading judgments from optical flow alone even when all points at differing depths are separated by at least 6 degrees in the visual field (Perrone & Stone, ARVO 1991; Stone & Perrone, Neurosci. Abst., 1991). This result is inconsistent with the local differential motion algorithm (Longuet-Higgins & Prazdny, Proc. R. Soc., 1980) which requires depth differences between closely spaced points. However, one could argue that this previous result might be explained by an enhanced algorithm (Rieger & Lawton, JOSA, 1985) which allows the use of points at greater separations at the expense of increasing both random and systematic errors in heading estimates. Methods. To examine this possibility, we devised a stimulus for which the Rieger-Lawton model predicts severe systematic errors. We measured observers' heading estimates in response to simulated forward motion (2 m/s) combined with yaw rotation (2 degrees/s left or right). The stimulus consisted of two half-planes of dots at different distances (12.5 m and 25 m), each occupying only the left or right half of the display (30 degrees H x 22 degrees V) with a 6 degrees wide central vertical gap separating the two half-planes. Results. The three observers (1 naive) were able to estimate heading under these conditions, although there was a moderate bias (mean: 4.5 degrees) in the direction of the rotation regardless of which half-plane was closer. In contrast, simulations of the Rieger-Lawton model predict a large bias (approximately 11 degrees) in the direction of the farther plane regardless of the direction of rotation. Conclusion. These results suggest that the local differential motion algorithm does not underlie human self-motion perception.
Publication Types:
Keywords:
- Algorithms
- Animals
- Head
- Humans
- Judgment
- Models, Biological
- Motion Perception
- Photic Stimulation
- Rotation
- Vision
- Visual Fields
- NASA Center ARC
- NASA Discipline Neuroscience
- NASA Discipline Space Human Factors
Other ID:
UI: 102233990
From Meeting Abstracts