GPS/INS and LiDAR Odometry Integration for Autonomous racecar

Описание: In the context of multi-modal systems, ensuring precise transformations between various global frames is crucial.
To effectively use a LiDAR scan-matched pose, as a multimodal measurement, generating a unified map based on GPS frames becomes an intuitive choice.
This is particularly true since the LiDAR scan-matched pose can be readily integrated.

To facilitate this process, we employ a technique known as pose-graph optimization. This method is tailored to construct a unified 3D map by minimizing the spatial disparities between corresponding poses within the map. Significantly, this section introduces a specialized approach for achieving frame unification through pose-graph optimization, with a specific focus on the racetrack environment.