Stand-alone obstacle avoidance controller for image-based navigation

• A. Durand-Petiteville UFPE
• V. Cadenat LAAS-CNRS / Université Paul Sabatier Toulouse III
• T. Sentenac LAAS-CNRS / Institut Clément Ader / IMT Mines Albi / INSA / UPS / ISAE

Resumo

In this paper we present a stand-alone reactive avoidance controller inspired by the spirals described by insects during their flights. Unlike state-of-the-art obstacle avoidance methods, it only focuses on circumventing hazardous elements and can be coupled to navigation controllers working in different state space, such as the image one. We describe the obstacle avoidance problem by following a set of spirals, the spiral parameters computation to deal with convex/concave obstacles, and the spiral following controller design. Finally, a set of simulations obtained with ROS/Gazebo show the efficiency of the proposed approach to navigate in a cluttered environment.

Palavras-chave: Measurement, Location awareness, Spirals, Navigation, Insects, Conferences, Computational modeling