The role of static scene information on locomotion distance estimation

Abstract

Judgement of travelled distance is important for many spatial learning and navigation tasks, including path integration. Research on visual information for locomotion distance estimation has traditionally focused on the effect of motion cues. However, static visual information is also generally available during real world navigation. This study investigated the contribution of static visual information on the estimation of travelled distance using a static-frame paradigm. Participants viewed successive static snapshots of simulated self-locomotion along a tunnel containing either random texture or geometric shapes, and then reproduced the travelled distance in a different environment. The results showed that when the motion information was removed, the use of actual distance information was eliminated in the texture condition but remained when perceptual landmarks were present, suggesting that static scene information from identifiable objects can support locomotion distance judgements. Moreover, travel duration was also used in conjunction with the spatial information. These findings suggest that studies involving locomotion distance estimation need to consider the potential contributions of the static scene mechanism and the temporal information in addition to the motion-based mechanism.

Keywords:

• Distance estimation
• Locomotion
• Optic flow
• Path integration
• Static scene

Notes

¹Not all path integration processes require locomotion distance estimation. For example, ants travelling along random, winding trajectories and performing path integration based on continuous vector integration (e.g., Müller & Wehner, 1988) do not require locomotion distance estimation.

²An integration of the velocity vector (including both speed and direction) over time yields a displacement vector (distance and direction) of the origin relative to an animal's current position. This is the process of path integration. An integration of the velocity value (speed alone) over time yields a distance value that represents the path length from the origin to the animal's current location. This path length is the locomotion distance (travelled distance). These two distance values are the same if the animal moves along a straight line, and are different if the trajectory is curved.

³For self-displacement judgements, the perceptual landmarks do not need to be unique. When multiple identical objects are present locally, this mechanism is useful only when the correspondence between successive instances can be determined.

⁴The change in egocentric distance reflects the linear displacement. For nonlinear paths, the trajectory is needed to judge the path distance. When the trajectory is unknown, the actual path length may be approximated by the summation of small segments of linear displacements.

⁵The perspective of the virtual scene was determined according to the position of the observer relative to the computer monitor, so that the computer display produced approximately the same retinal image on the observer as if s/he was watching a tunnel through a window. Thus, the absolute egocentric distance of an object in the scene can in principle be determined as in real environments with comparable information.