The visual control of walking over complex terrain with flat versus raised obstacles

摘要

When humans walk over terrain with irregularly spaced flat obstacles, such as puddles or patches of ice, visual information about the locations of those obstacles is needed at least two step lengths in advance of foot placement to avoid collisions and energetically costly adjustments to the natural gait cycle (Matthis and Fajen, 2013, 2014). However, obstacles found in natural environments are not always flat and the height of an obstacle, as well as its location, constrains how the feet must be moved to successfully avoid a collision. The aim of the present study was to determine how the presence of raised obstacles affects how far ahead walkers need to see to avoid collisions and maintain walking speed and energetic efficiency. Subjects traversed a short path over virtual complex terrain comprising irregularly spaced obstacles. The virtual scene was projected onto the floor using a 3D-capable projector that rendered a stereoscopic image at 60fps per eye. When viewed while wearing LCD shutter glasses, subjects experienced the illusion of terrain with variations in surface elevation. This paradigm was developed and validated by Diaz, et al. (2015) as a means of providing realistic, stereoscopic visual information without an HMD and allowing for natural obstacle avoidance behavior. To evaluate when visual information is necessary for successful locomotion, we created a traveling window of visibility centered on the subject's head and made obstacles visible only when they fell within the window. The key manipulations were the size of the visibility window, which ranged from 1.0 to 5.0 step lengths, and whether the obstacles were 3D or flat. The results reveal how variations in the structure of the environment affect how far in advance walkers begin to use visual information about the upcoming terrain as well as their ability to exploit their inverted-pendulum-like structure.