A framework for RF-Visual SLAM

Simultaneous Localization and Mapping, SLAM, is an important topic in the field of robotics and autonomous navigation. The metric SLAM suffers from sensor inaccuracies and thus cannot be used for long-term navigation. In such case, Visual SLAM or a Hybrid SLAM based on both metric and visual approach is a good alternative. In this paper, in order to speed up a Visual SLAM, we propose a novel concept of dynamic dictionary generated on the results of triangulation done on RF, radio frequency, signals from nearest cell towers of a cellular network. This dynamic dictionary efficiently manages the scalability of a Visual SLAM and make it possible to work in a large-scale environment. A framework is proposed along with triangulation data of a city and with simulations to support the concept.