An efficient grid based navigation of wheeled mobile robots based on visual perception

Abstract

Navigation is one of the major challenges and a key ability for the mobile robots to accomplish the given task. It is a two point problem where the accurate positioning and path planning is necessary. The efficient path planning and positioning with the use of conventional sensors becomes cumbersome when the robots are resource constrained and uses minimal information of the environment. Therefore, a simple heuristic navigation algorithm based on grids is addressed in this paper. A simple visual perception method with a use of line scan camera is also employed in this work, as an alternative to the existing methods with conventional sensors, towards the identification of grids and lines in the workspace. The efficiency of the proposed technique is investigated experimentally by applying on a miniature wheeled mobile robot "Grid-bot" and the results are discussed. The results reveals that the proposed method be a viable alternative for a resource constrained robot to plan its path and navigate the environment efficiently in indoor applications. © 2012 Springer-Verlag.