Visual Landmark Navigation Through Large-scale Environments

Several theoretical models have been proposed for the visual homing behaviour of insects. We take one such model for visual homing, and extend it to an algorithm capable of autonomous exploration and navigation through large-scale environments. The algorithm uses a novel approach to waypoint selection during the construction of multi-leg routes. By locating the boundaries between visual locales, and placing waypoints at these edges, advantages are gained over current biomimetic models of similar competence. Although ultimately, the algorithm is meant for implementation on autonomous mobile robotics, direction in design is taken from biological observation with the aim the algorithm will remain biologically plausible.