This is a repository that includes all the codes to geo_correct 3d point clouds generated by the gantry 3d laser scanner. Before running this repository one should merge east and west point clouds using [https://github.com/phytooracle/3d_icp_point_cloud_registration] repository and then properly rotate and register the merged ply file using [https://github.com/phytooracle/3d_geo_registration] repo. The docker file can be used to create a container image (or a singularity image) and use it to run (exec) the geo_correct_point_cloud.py script. This repository uses the detected locations of the plants from RGB sensor to geo-correct the point clouds. The output of this repository is the geo-corrected ply file.
- Required Arguments:
- The path to the registered and rotated point cloud file: path (-p --path)
- Path to the detecto model pth file: model (-m --model)
- The path to the full plant detection file: plants (-d --plants)
- The date of scan, used for picking the closest RGB detection: scandate (-s --scandate)
- The output directory: output (-o --output)
- Optional Arguments:
- Number of cores to use: core (-c --core)
- Docker:
- Docker run [-v /mount_this:/to_this] image_name python3 geo_correct_point_cloud.py [params]
- Singularity:
- Singularity exec [-B /mount_this:/to_this] image_name.simg python3 geo_correct_point_cloud.py [params]