Topological mapping for mobile robots using a combination of sonar and vision sensing

摘要: Topological maps represent the world as a network of nodes and arcs: are distinctive places in environment arcs paths between places. A significant issue building topological is defining Most previous work mapping has concentrated on using sonar sensors to define However, limited range angular resolution, which can make it difficult distinguish different Our approach combines sonar-based definition with visual information. We use robot's determine where capture images cues extracted from those help perform place recognition. Information these two sensing modalities combined simple Bayesian network. Results described this paper show that our robot able recognition without having move through sequence places, case most currently implemented systems.