Validate Livox scanner using real point cloud

[han16nah]

Problem

While we get reasonable results visually, we never validated the risley deflector using real point clouds. We should especially have a look at the scan pattern, point density depending on the integration time, and the field of view.

Solution

Livox provides point cloud data recorded with the Livox mid-40 (in spherical coordinates) with an integration time of 4s: https://terra-1-g.djicdn.com/65c028cd298f4669a7f0e40e50ba1131/Download/update/Mid-40%20Point%20Cloud%20Explanatory%20Notes.zip

A notebook for comparison of real and simulated point clouds is available in the risley branch: example_notebooks/8_TLS_livox_demo_testing.ipynb Here, it is found that the simulated point cloud has a larger field of view...

A further question arises: The rotation speeds of the prisms are taken from Liu et al. 2022 (see V. Conclusion, Section A). Here, they are provided in rotations per minute, while we use them as Hz. Is this an issue/does it have to be converted? Or do we just have to rename our attribute names (rotorFreq1_Hz and rotorFreq2_Hz)?