Mobile robot localization and path planning based on Feature recognition and Dijkastra's Algorithm

This paper proposes a fast localization method, based on a simple feature recognition like a sphere. The local and the global position is obtained by scenario reconstruction, and feature recognition as the reference for localization. A 2D laser range finder mounted on a tilting mechanism scans the 3D information of the environment. The 3D workspace is divided by triangulation, and a path planned for fast obstacle avoidance created by connecting the center points of each of the generated tetrahedrons using Dijkstra’s algorithm.