Abstract
When an observer moves straight through a stationary scene, the images of objects move in a radial pattern. One can detect a moving object in the scene if its image moves in a different direction from this pattern. However, a speed difference is ambiguous. It could be due to distance variation or to object motion. We examined how the angle or speed of object motion influences whether people perceive the scene as rigid or non-rigid. Observers viewed a 38×38 deg window within which a field of 500 random dots moved to simulate observer motion toward two transparent planes. The object consisted of a 10×10 deg square of 30 dots initially located 8 deg to the right of center. To test the effect of object angle we kept the object speed constant and varied the angle of object motion between 0 and 360 deg. To test speed, we kept the angle constant and varied the speed by a factor between 0.25 to 2.75 times the speed of dots in the nearest plane. Subjects indicated by a keypress whether the scene appeared rigid or non-rigid. Results for 5 observers show that perception of rigidity decreased from 90% rigid responses to 30% as the angle increased from 0 to 157.5 deg. Perceived rigidity rose slightly for an angle of 180 deg. Perception of rigidity also fell as the speed factor deviated from 1.0, although these displays were all consistent with a rigid interpretation, with the object placed at different distances from the observer. Speed factors of .25, 1.0 and 2.75 resulted in 38%, 98% and 4% rigid responses, respectively. Thus, in the absence of other cues, a moving observer may perceive an object located sufficiently closer or farther from the other items in the scene as a moving object.
Supported by NSF grant #IBN-0343825.