Event
Proposal of the full 6-DOF sonar SLAM system RTISSLAM-3D
1 reports · 1 sources
Jan Steckel, Vincent Katter, Wouter Jansen, and others proposed RTISSLAM-3D, a 6-DOF SLAM system for wheeled robots equipped with an aerial 3D imaging sonar, an IMU, and wheel encoders. The system uses a spherical CFAR detector to convert sonar images into sparse 3D detections, and combines 3D acoustic odometry that models the sonar beam shape, place recognition using polar submap descriptors, and loop-closure verification via 3D registration to form a 6-DOF pose graph with pose factors. Evaluation over a 545-meter drive in a forest (with 3D laser SLAM as reference) showed that loop closure reduced the horizontal error of acoustic odometry from 1.74 meters to 0.28 meters, with a height error of 0.12 meters; 94.6% of the generated 3D sonar map lies within 0.3 meters of the laser map.
Report timeline
- RTISSLAM-3D: A Full 6-DoF SLAM System Based on an In-Air 3D Imaging Sonar
The paper proposes RTISSLAM-3D, a 6-DoF SLAM system for a wheeled robot equipped with an in-air 3D imaging sonar, an IMU, and wheel encoders. The system uses a spherical CFAR detector to reduce sonar images to sparse 3D detections, which are used for 3D acoustic odometry that models the sonar beam shape, for place recognition with polar submap descriptors, and for loop-closure verification via 3D registration, all combined into a 6-DoF pose graph with attitude factors.
arXiv cs.RO ·