Autonomous Implicit Indoor Scene Reconstruction with Frontier Exploration
ICRA 2024(2024)
Abstract
Implicit neural representations have demonstrated significant promise for 3Dscene reconstruction. Recent works have extended their applications toautonomous implicit reconstruction through the Next Best View (NBV) basedmethod. However, the NBV method cannot guarantee complete scene coverage andoften necessitates extensive viewpoint sampling, particularly in complexscenes. In the paper, we propose to 1) incorporate frontier-based explorationtasks for global coverage with implicit surface uncertainty-basedreconstruction tasks to achieve high-quality reconstruction. and 2) introduce amethod to achieve implicit surface uncertainty using color uncertainty, whichreduces the time needed for view selection. Further with these two tasks, wepropose an adaptive strategy for switching modes in view path planning, toreduce time and maintain superior reconstruction quality. Our method exhibitsthe highest reconstruction quality among all planning methods and superiorplanning efficiency in methods involving reconstruction tasks. We deploy ourmethod on a UAV and the results show that our method can plan multi-task viewsand reconstruct a scene with high quality.
MoreTranslated text
Key words
View Planning for SLAM,Mapping,Continual Learning
AI Read Science
Must-Reading Tree
Example
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
Generate MRT to find the research sequence of this paper
Chat Paper
Summary is being generated by the instructions you defined