A mapping of visual SLAM algorithms and their applications in augmented reality

Abstract

The term visual SLAM defines the problem of build a map of an environment and perform location, simultaneously. In the augmented reality experience, we can apply SLAM techniques to insert virtual elements in the user's real-world view according to their observation point (location) and environment structure (mapping). In this work, we investigated the main algorithms of visual SLAM, and its applications in augmented reality. Here, we describe the key features of these algorithms and two taxonomies for SLAM techniques are proposed.