VDBblox: Accurate and Efficient Distance Fields for Path Planning and Mesh Reconstruction
Yinlong Bai,Zhiqiang Miao,Xiangke Wang,Yong Liu,Hesheng Wang,Yaonan Wang
DOI: https://doi.org/10.1109/IROS55552.2023.10342123
2023-01-01
Abstract:Highly accurate and efficient map in unknown and complex environments is essential for robotics navigation. Traditionally, mobile robot platforms are often computationally constrained when using multiple sensors to process large amounts of input data. In previous works, some of them have been deployed to embedded platforms in real-time. However, how to balance accuracy and efficiency while reducing the computational resources and the memory footprint is still the bottleneck. Motivated by these challenges, we proposed a mapping framework called VDBblox to incrementally build Euclidean Signed Distance Fields (ESDFs) map from Truncated Signed Distance Fields (TSDFs) mapping. We use a novel weight function to update the non-projective TSDFs, thus improving the quality of the mesh reconstruction with higher accuracy than up-to-date methods. Meanwhile, the generated ESDFs map is maintained by the least recently used (LRU) cache to dynamically handle the obstacle changes with less runtime than state-of-the-art. We show VDBblox performance in terms of accuracy and efficiency by benchmark comparison on RGB-D and LiDAR public datasets. Moreover, we demonstrate that VDBblox can be integrated into a completed quadrotor system as a sub-module. Then we validate it through online obstacle avoidance and high-quality mesh reconstruction in real-world experiments. Finally, we release our method as open-source code to the community(1).