Volumetric Data Fusion of External Depth and Onboard Proximity Data For Occluded Space Reduction
This work addresses the issue of occluded spaces in robotic mapping, which is incremental as it builds on the Octomap framework.
The paper tackles the problem of mapping a robot's environment by probabilistically fusing external depth and onboard proximity data to create a volumetric 3-D map, resulting in a more accurate map with fewer occlusions as shown in simulated results.
In this work, we present a method for a probabilistic fusion of external depth and onboard proximity data to form a volumetric 3-D map of a robot's environment. We extend the Octomap framework to update a representation of the area around the robot, dependent on each sensor's optimal range of operation. Areas otherwise occluded from an external view are sensed with onboard sensors to construct a more comprehensive map of a robot's nearby space. Our simulated results show that a more accurate map with less occlusions can be generated by fusing external depth and onboard proximity data.