Topological Map Merging

Multi-robot Topological Exploration Using Olfactory Cues

“…This is a difficult problem when the robots do not have a common reference frame or global positioning system [11]. Topological maps provide a brief characterization of the navigability of an environment, and, with measurements easily collected during exploration, the vertices of the map can be embedded in a metric space [11].…”

“…This is a difficult problem when the robots do not have a common reference frame or global positioning system [11]. Topological maps provide a brief characterization of the navigability of an environment, and, with measurements easily collected during exploration, the vertices of the map can be embedded in a metric space [11]. These maps use a graph to represent possibilities for navigation through an environment.The current proposed approach employs a topological mapping technique, so the robots only exchange a few environmental features.…”

“…Using topological maps, the problem of "map merging" is reduced to the "graph merging" problem that is already addressed in several ways [11,12,13]. Whereas most approaches to topological map merging and related problems have focused on using either map structure [14] or map geometry [15], our algorithm takes advantage of both.…”

“…Whereas most approaches to topological map merging and related problems have focused on using either map structure [14] or map geometry [15], our algorithm takes advantage of both. Similar to [11], the use of map structure allows quick identification of potential vertex matches in the maps (and rejection of mismatches), while the use of map geometry enables the algorithm to directly merge maps with multiple (disconnected) overlapping regions. In another words, the algorithm in [11] uses both the structure and the geometry of topological maps to determine the best correspondence between maps with single or multiple overlapping regions.…”

“…Similar to [11], the use of map structure allows quick identification of potential vertex matches in the maps (and rejection of mismatches), while the use of map geometry enables the algorithm to directly merge maps with multiple (disconnected) overlapping regions. In another words, the algorithm in [11] uses both the structure and the geometry of topological maps to determine the best correspondence between maps with single or multiple overlapping regions. However, if the geometric data of the local maps are not obtained through a unique coordinate system this method (in [11]) will not be functional.…”