Path planning in image space for the autonomous navigation of unmanned vehicles in unstructured outdoor environments

An approach to stereo based local path planning in unstructured environments is presented. The approach differs from previous stereo based and image based planning systems (i.e. top-down occupancy grid planners, autonomous highway driving algorithms, and view-sequenced route representation), in that it uses specialized cost functions to find paths through an occupancy grid representation of the world directly in the image plane, and forgoes the standard projection of cost information from the image plane down onto a top-down 2D Cartesian cost map. Three cost metrics for path selection in image space are discussed. A basic image based planning system is presented, and its susceptibility to rotational and translational oscillation is discussed. Two extensions to the basic system are presented that overcome these limitations---a cylindrical based image system and a hierarchical planning system. All three systems are implemented in an autonomous robot and are tested against a standard top-down 2D Cartesian planning system on three outdoor courses of varying difficulty. It was found that the basic image based planning system fails under certain conditions; however, the cylindrical based system is well suited to the task of local path planning and for use as a high resolution local planning component of a hierarchical planning system.