Vision SLAM Maps: Towards Richer Content

摘要: Simultaneous Localization and Mapping (SLAM) is a recursive probabilistic inferencing process for concurrently building map of robot’s surroundings localizing that robot within this map. The ultimate goal SLAM to operate anywhere, allowing navigate autonomously producing meaningful purposeful Research in date has focused on improving the localization part SLAM, while lagging ability produce useful maps. Indeed, all feature-based maps are built from either low level features such as points or lines artificial beacons; have little use other than perform SLAM. There benefits real natural objects indigenous environment operations surveying remote areas guide human navigation dangerous settings. To investigate potential maps, an Inertial-Visual system designed used here which relies inertial measurements predict ego-motion digital camera collect images landmarks about scene. Experiments conducted mobile vehicle show encouraging results highlight Vision generate agree with ground truth. Computer capable recognizing type, detecting trees environment, different based clusters distinctive visual features.