pymccrgb is a Python package for multiscale curvature classification of point clouds with color attributes.
It extends a popular classification method (MCC lidar) [0] to point cloud datasets with multiple color channels, such as those commonly produced in surveys using drone photography or other platforms. It can be used to distinguish points from the ground surface and low vegetation in data produced by structure from motion photogrammetry, stereo photogrammetry, or multi-spectral lidar scanning, or to filter colorized lidar point clouds in LAS/LAZ or CSV format.
The intended users are scientists in geomorphology, ecology, or planetary science who want to classify point clouds for topographic analysis, canopy height measurements, or other spectral classification.
This package is developed for Linux and Python 3.6+. It depends on common Python packages like sklearn, numpy, the LibLAS C API, and MCC Python bindings.
You can install it with conda
or virtualenv
in a virtual environment.
git clone https://github.com/rmsare/pymccrgb
cd pymccrgb
conda env create -f environment.yml
conda activate pymcc
pip install pymccrgb
py.test pymccrgb/tests
(Conda package coming soon)
The LibLAS C library is required for MCC and pymccrgb
. The MCC wrapper also
requires Boost and the C++11 or later standard library. These are installed
with the conda package.
Refer to the documentation and the LibLAS install guide for instructions for installing LibLAS from source.
Example notebooks are available in the docs or at docs/source/examples.
from pymccrgb import mcc, mcc_rgb
from pymccrgb.datasets import load_mammoth_lidar
from pymccrgb.plotting import plot_results
# Load sample data (Mammoth Mountain, CA)
data = load_mammoth_lidar(npoints=1e6)
# MCC algorithm
ground_mcc, labels_mcc = mcc(data)
# MCC-RGB algorithm
ground_mccrgb, labels_mccrgb = mcc_rgb(data)
plot_results(data, labels_mcc, labels_mccrgb)
Results of MCC and MCC-RGB on a forested area near Mammoth Mountain, CA.
Read the documentation for example use cases, an API reference, and more at pymccrgb.readthedocs.io.
Bug reports are much appreciated. Please open an issue with the bug
label,
and provide a minimal example illustrating the problem.
Feel free to suggest new features in an issue with the new-feature
label.
If you would like to add a feature or fix a bug, please fork the repository, create a feature branch, and submit a PR and reference any relevant issues. There are nice guides to contributing with GitHub here and here. Please include tests where appropriate and check that the test suite passes (a Travis build or pytest pymccrgb/tests
) before submitting.
Please open an issue with your question.
[0] Evans, J. S., & Hudak, A. T. 2007. A multiscale curvature algorithm for classifying discrete return LiDAR in forested environments. IEEE Transactions on Geoscience and Remote Sensing, 45(4), 1029-1038 doi
This work is licensed under the MIT License (see LICENSE). It also
incorporates a wrapper for the mcc-lidar
implementation,
which is distributed under the Apache license (see LICENSE.txt).