Visual based Localization for mobile robots with Support Vector Machines

Abstract

Support vector machine (SVM) is a relative new classification algorithm with some advantages over other machine learning methods. This paper presents a SVM based localisation algorithm for indoor robot navigation by recognizing the environmental features that are known as priori. A topological map is adopted by the mobile robot for the coarse global localisation. Then at each topological node located, geometrical grids are adopted for the proposed algorithm to provide fine position information of the mobile robot. Experimental results are presented to show the feasibility and good performance of the proposed method