Development of Embedded Algorithm for Visual Simultaneous Localization and Mapping

Abstract

The Simultaneous Localization and Mapping (SLAM) problem is recurrent in today's robotics. One challenge of it is the extensive computational cost to create complex maps in real-time. Various applications, mainly search and rescue operate in GPS denied scenarios, with possible difficulty communicating with an external base. A portable SLAM system capable of being run in a microcomputer would greatly help such operations. This paper mentions the unfinished into this topic and discusses further steps that shall be taken in the upcoming months.