Autonomous UAV Flight and Mapping beneath Forest Canopy (high-fidelity simulated and real forests)

[@MrMax836]

Hello Please tell me how you determine the floor (ground) in an image? Segmentation? How is it then determined on the map at what height the voxel should be placed?

[@robinvishen4697]

Great project and implementation. Is there any specific reason for using Unity? And what's the microprocessor used? I am working on a similar project on UGV, so I'd really appreciate if you could answer these.

[@samliu1726]

A major reason for using Unity is that it offers high-fidelity visual inputs, which are important not only for our vision-based state estimation, but also for our ongoing research in extracting semantics from the environment. For field experiments, we use Intel NUC (processor: i7-8559U) as our onboard computer.

[@ujasmandavia8481]

Do you have a paper published for it? I would love to read it.

[@samliu1726]

For our autonomous flight system, please refer to our lab's previous paper Fast, autonomous flight in GPS‐denied and cluttered environments by Kartik Mohta et al. (onlinelibrary.wiley.com/doi/abs/10.1002/rob.21774). For SLOAM, please refer to our paper Sloam: Semantic lidar odometry and mapping for forest inventory by Steven Chen et al (ieeexplore.ieee.org/abstract/document/8949363).