Real-time Semantic 3D Dense Occupancy Mapping with Efficient Free Space Representations
This addresses the bottleneck of mapping speed for real-world deployments in robotics or autonomous systems, though it appears incremental as it builds on existing Bayesian kernel inference.
The paper tackles real-time semantic 3D occupancy mapping by proposing two novel free space representations to improve speed, achieving real-time performance on a consumer-grade CPU and handling dynamic scenarios.
A real-time semantic 3D occupancy mapping framework is proposed in this paper. The mapping framework is based on the Bayesian kernel inference strategy from the literature. Two novel free space representations are proposed to efficiently construct training data and improve the mapping speed, which is a major bottleneck for real-world deployments. Our method achieves real-time mapping even on a consumer-grade CPU. Another important benefit is that our method can handle dynamic scenarios, thanks to the coverage completeness of the proposed algorithm. Experiments on real-world point cloud scan datasets are presented.