Sensor-based multisource fusion in an unstructured environment mapping and localization

For the complex test environment and uneven road surface, a sensor multisource fusion framework was used for point cloud map construction and localization. The internal parameters of IMU were first calibrated, and then the spatial parameters of LiDAR and IMU were calibrated. The calibrated multi-sensor fusion algorithm has better map building effect and stronger localization robustness, which is valuable for mobile robot localization and navigation applications, through experiments at Taian Jichuang robot test site.