Plane Detection with Feature Point Tracking from Monocular Images on a Mobile Robot for Indoor Environments Oscar Alonso-Ramirez

Abstract—In this paper we propose a methodology for robust plane recovering from monocular images. The method is mainly based on the tracking interest points taken from a mobile robot’s camera. An efficient methodology for interest point detection, tracking and recovering has been implemented. The method proposed works in real time, and it is robust enough to deal with different undesired camera motions as the presented with legged robots. Two mobile robotic platforms were used to validate results a wheeled differential robot and a humanoid robot (with legs)