Mobile Robot Visual Navigation Systems using Optical Flow

This paper aims to introduce navigation frameworks that rely on visual perception for a mobile robot equipped with wheels, executed through optical flow methodologies. Two discriminant algorithms are investigated to analyze visual cues in unfamiliar static environments. The control systems proposed rely on estimations of movement and decision-making mechanisms that use the measured data calculated by the optical inflow algorithms to direct the robot autonomously within its working area. The results are analyzed and explored, to showcase the capacity for unimpeded movement by a mobile robot using visually-based navigation systems, without any collisions with obstructions.