A method for indoor and outdoor collaborative localization and mapping based on multi-sensors unmanned platforms

Abstract

This paper focuses on target detection tasks in complex environments and explores a heterogeneous unmanned node cross-scenario precise detection method. Specifically, the paper unifies the location information detected by the platform entering the indoor environment to the local geographic frame where the swarm is located. In our design, a simple collaborative detection is achieved using a quadruped robot equipped with a laser radar for mapping and detection with its flexible mobility to enter the indoor environment, and a wheeled unmanned vehicle carrying high-precision IMU and GNSS to provide positioning benchmarks with its excellent load-carrying performance. The relative positions between unmanned platforms are calculated through UWB and depth camera, and the coordinates of the quadruped robot in the local geographic frame are successfully solved, thus inferring the location information of the indoor environment detected by the robot in the local geographic frame.