Robot Localization
Overview
Robot localization estimates a robot’s pose over time, either relative to its starting point or within a global map, so it can navigate and act safely. Common approaches fuse cameras, LiDAR, radar, inertial sensors, GNSS, and radio ranging with Kalman filters, particle filters, or factor-graph optimization; map matching and place recognition help recover pose or recognize previously visited places. Recent methods also use semantic landmarks, cross-view imagery, learned representations, and cooperative measurements shared between robots. A central challenge is maintaining reliable, well-calibrated estimates when sensors lose information, environments change, or scenes are repetitive, while keeping computation practical for onboard use.
Latest papers
179 papers21 papers in the last four weeks, up 11% on the four weeks before. 0.1% of all new papers.