A repository for pointcloud pre- and post- processing scripts using a variety of python tools.

Chief among them is the python cloudCompare wrapper "CloudComPy310", which provides several cloud cleaning and analysis routines.

You can run the point cloud contraction and centreline extraction routine using the compute_centrelines.ipynb notebook. The main idea is to iteratively contract the set of points to generate a zero-volume approximation of the curve skeleton of the cave conduit and eventually construct a 3D polyline that describes the conduit in a more general way. Part of the process includes converting the centreline in ASCII format to other interoperable formats, namely the AutoCAD DXF format, as well as the geographic JSON format.

We provide an example notebook showcasing some of the CloudComPy library routines, namely for running the Cloth Simulation Filter to segment the floor from the conduit ceiling and also computing the Illuminance value (visible sky portion, PCV).

You can run the rasterisation routine using the extract_rasters.ipynb notebook. Rasterisation is a process turning the 3D data set of point positions to a 2.5D image, containing for each pair of x and y coordinates a single elevation value. Raster images can be post processed in any GIS software or dedicated code libraries. Here we present the routine used to rasterise conduit floor and ceiling.

You can find a binary for CloudComPy*_-date-.7z here and alternatively, building instructions here. Unpack this pre-compiled binary in a repository of your choice.

You can create a new conda environment using the environment.yml file given in this repository. To do this, open a command prompt, navigate to the pc-processing repository and run the following command:

conda env create -f environment.yml

To launch a kernel with the PCProcessing environment, you can launch a command prompt window, navigate to the pc-processing repository and run the following command to activate the environment:

conda activate PCProcessing

Then this command to check that the module is installed properly and all checks pass.

<path-to-CloudComPy310-binary>/envCloudComPy.bat

Finally, launch a jupyter kernel with either of the following:

jupyter notebook

jupyter server

jupyter lab