Vision-based unsupervised learning of unexplored environment for autonomous land vehicle navigation

[[abstract]]In this paper, the authors present a vision-based approach, designed for autonomous land vehicle navigation, which is used for unsupervised learning of the environment. The learning system is comprised of three subsystems: feature location, model management and environment exploration. These three strategies are described. Simulation and experimental results indicate the feasibility of using this proposed approach for a learning system for real environments