A Real‐Time and Fast LiDAR–IMU–GNSS SLAM System with Point Cloud Semantic Graph Descriptor Loop‐Closure Detection

Herein, a real-time, fast, tightly coupled simultaneous localization and mapping (SLAM) system is proposed. The deep neural network used to segment the point cloud semantically construct semantic map descriptor, global navigation satellite real-time kinematic position detect loop closure. Finally, factor-graph model for optimization. working principle of SLAM introduced in detail, including construction factor graph, generation graph ring loop-closure detection process, Indoor outdoor experiments are conducted validate performance overall trajectory estimation detection. Compared traditional methods, proposed method offers improved accuracy estimation. It effectively addresses issues such as limited feature constraints, large computational consumption, subpar performance, allowing quick accurate Moreover, demonstrates time consumption reduction two-thirds compared approaches while exhibiting superior performance.