SceneSLAM: A SLAM framework combined with scene detection

Abstract

Different sensors have different capacities and may even fail in different scenes. This will influence accuracy and reliability of SLAM systems utilizing those sensors. To enhance the robustness of SLAM systems, we present SceneSLAM in this paper, which is a novel extensible SLAM framework by combing different SLAM systems facilitated by a scene detection method. SceneSLAM can activate different SLAM modules automatically based on the result of scene detection and perform map fusion to achieve globally consistent localization and mapping. To verify the extendibility and effectiveness of SceneSLAM, we build a prototype system based on SceneSLAM to enhance the reliability of existing SLAM systems when the light condition changes dramatically. The prototype system runs on a TurtleBot robot equipped with a Kinect sensor. The experimental results show that SceneSLAM provides an effective solution to enhance the robustness of existing SLAM systems in dynamic environments by activating corresponding SLAM module based on the scene detection results. SceneSLAM provides efficient support to developing and implementation of a SLAM system with multiple SLAM modules and a scene detection module. Furthermore, we release the source code of our framework with our prototype system, so that it can be easily extended by other researchers.