A coarse-to-fine approach for fast path finding for mobile robots

Abstract

A simple but effective method for fast path finding for mobile robots in grid-based search space is presented. The paper presents a systematic way to reduce the resolution of a grid map preserving the occupancy structure of the original map. Paths are found in two steps with varying quantization levels of the space in coarse-to-fine manner. The resulting paths are practically optimal and experimental results show a great reduction of the searching time, making it possible to implement a real-time global path planner in a very large environment.