Vision Based Simultaneous Localization and Mapping

Abstract

This paper gives an introduction to the Simultaneous Localization and Mapping (SLAM) methods and recent advances in computational methods of SLAM problem for large scale & complex environment. SLAM addresses the problem of a robot navigating through an unknown environment. It is a process by which a mobile robot can build a map of its near-by environment and at the same time use this map to compute its own location. There are number of approaches to the SLAM problem. Different algorithms have been used to perform SLAM including Extended Kalman Filtering, Particle Filtering, Local Bundle Adjustment etc. There are still many practical issues to overcome, especially in more complex outdoor environments. This present paper elucidates variants of the SLAM problem and proposes a taxonomy for the same.