Browsing by Subject "Simultaneous Localization and Mapping"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    A distributed framework for monocular visual SLAM
    (2017-04) Egoda Gamage, Ruwan; Tuceryan, Mihran; Computer and Information Science, School of Science
    In Distributed Simultaneous Localization and Mapping (SLAM), multiple agents generate a global map of the environment while each performing its local SLAM operation. One of the main challenges is to identify overlapping maps, especially when agents do not know their relative starting positions. In this paper we are introducing a distributed framework which uses an appearance based method to identify map overlaps. Our framework generates a global semi-dense map using multiple monocular visual SLAM agents, each localizing itself in this map.