FIT-SLAM -- Fisher Information and Traversability estimation-based Active SLAM for exploration in 3D environments

• DOI: 10.48550/arxiv.2401.09322
• Publication Date: 2024-01-01
• Authors: Suchetan Saravanan; Corentin Chauffaut; Caroline Ponzoni ...; Damien Vivet

Active visual SLAM finds a wide array of applications in GNSS-Denied sub-terrain environments and outdoor for ground robots. To achieve robust localization mapping accuracy, it is imperative to incorporate the perception considerations goal selection path planning towards during an exploration mission. Through this work, we propose FIT-SLAM (Fisher Information Traversability estimation-based SLAM), new method tailored unmanned vehicles (UGVs) explore 3D environments. This approach devised with dual objectives sustaining efficient rate while optimizing accuracy. Initially, estimation global traversability map conducted, which accounts environmental constraints pertaining traversability. Subsequently, candidate along that takes into account information provided by landmarks used backend successful execution . The entire algorithm tested evaluated first simulated world, followed real-world environment compared pre-existing methods. results obtained evaluation demonstrate significant increase effectively minimizing covariance.