| Name | dendromatics JSON |
| Version |
0.4.2
JSON |
| download |
| home_page | |
| Summary | Automatic dendrometry and forest inventory for terrestrial point clouds |
| upload_time | 2024-03-15 07:59:27 |
| maintainer | |
| docs_url | None |
| author | |
| requires_python | >=3.8 |
| license | |
| keywords |
forestry
lidar
tls
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
.. image:: https://github.com/3DFin/dendromatics/blob/main/docs/_static/dendromatics_logo.png
:width: 300
:align: center
Dendromatics
============
|docs| |pypi| |tests|
.. |docs| image:: https://readthedocs.org/projects/dendromatics/badge/?version=latest
:alt: Documentation Status
:scale: 100%
:target: https://dendromatics.readthedocs.io/en/latest/?badge=latest
.. |pypi| image:: https://badge.fury.io/py/dendromatics.svg
:alt: PyPI status
:scale: 100%
:target: https://badge.fury.io/py/dendromatics
.. |tests| image:: https://github.com/3DFin/dendromatics/actions/workflows/test.yml/badge.svg
:alt: Tests status
:scale: 100%
:target: https://github.com/3DFin/dendromatics/actions
Description
===========
The *src* folder contains functionalities to detect the trees present in a terrestrial 3D point cloud from a forest plot, and compute individual tree parameters: tree height, tree location, diameters along the stem (including DBH), and stem axis. These are based on an updated version of the algorithm proposed by (Cabo et al., 2018).
The functionalities may be divided in four main steps:
0. Height-normalization of the point cloud.
1. Identification of stems among user-provided stripe.
2. Tree individualization based on point-to-stems distances.
3. Robust computation of stems diameter at different section heights.
Although individual, somewhat independent functions are provided, they are designed to be used in a script that calls one after the other following the algorithm structure. An example script can be found in `example` folder.
Examples
========
Height-normalization
--------------------
Almost all functions in the module expect a height-normalized point cloud to work as intended. If your point cloud is not height-normalized, you can do it in a simple way using some of the module functions. I'ts based on 'Cloth simulation Filter' (Zhang et al., 2016).
.. code-block:: python
import laspy
import numpy as np
import dendromatics as dm
# Reading the point cloud
filename_las = 'example_data.las' # your .las file
entr = laspy.read(filename_las)
coords = np.vstack((entr.x, entr.y, entr.z)).transpose()
# Normalizing the point cloud
dtm = dm.generate_dtm(clean_points)
z0_values = dm.normalize_heights(coords, dtm)
# adding the normalized heights to the point cloud
coords = np.append(coords, np.expand_dims(z0_values, axis = 1), 1)
If the point cloud is noisy, you might want to denoise it first before generating the DTM.
.. code-block:: python
clean_points = dm.clean_ground(coords)
Identifying stems from a stripe
-------------------------------
Simply provide a stripe (from a height-normalized point cloud) as follows to iteratively 'peel off' the stems.
.. code-block:: python
# Defining the stripe
lower_limit = 0.5
upper_limit = 2.5
stripe = coords[(coords[:, 3] > lower_limit) & (coords[:, 3] < upper_limit), 0:4]
stripe_stems = dm.verticality_clustering(stripe, n_iter = 2)
Individualizing trees
---------------------
Once the stems have been identified in the stripe, they can be used to individualize the trees in the point cloud.
.. code-block:: python
assigned_cloud, tree_vector, tree_heights = dm.individualize_trees(coords, stripe_stems)
Computing sections along the stems
----------------------------------
compute_sections() can be used to compute sections along the stems of the individualized trees.
.. code-block:: python
# Preprocessing: reducing the point cloud size by keeping only the points that are closer than some radius (expected_R) to the tree axes
# and those that are whithin the lowest section (min_h) and the uppest section (max_h) to be computed.
expected_R = 0.5
min_h = 0.5
max_h = 25
section_width = 0.02
xyz0_coords = assigned_cloud[(assigned_cloud[:, 5] < expected_R) & (assigned_cloud[:, 3] > min_h) & (assigned_cloud[:,3] < max_h + section_width), :]
stems = dm.verticality_clustering(xyz0_coords, n_iter = 2)[:, 0:6]
# Computing the sections
section_len = 0.2
sections = np.arange(min_h, max_h, section_len) # Range of uniformly spaced values within the specified interval
X_c, Y_c, R, check_circle, second_time, sector_perct, n_points_in = dm.compute_sections(stems, sections)
Tilt detection
--------------
tilt_detection() computes an 'outlier probability' for each section based on its tilting relative to neighbour sections and relative to the tree axis.
.. code-block:: python
outlier_prob = dm.tilt_detection(X_c, Y_c, R, sections)
For further examples and more thorough explanations, please check *example.py* script in */examples* folder.
References
==========
Cabo, C., Ordóñez, C., López-Sánchez, C. A., & Armesto, J. (2018). Automatic dendrometry: Tree detection, tree height and diameter estimation using terrestrial laser scanning. International Journal of Applied Earth Observation and Geoinformation, 69, 164–174. https://doi.org/10.1016/j.jag.2018.01.011
Prendes, C., Cabo, C., Ordoñez, C., Majada, J., & Canga, E. (2021). An algorithm for the automatic parametrization of wood volume equations from Terrestrial Laser Scanning point clouds: application in Pinus pinaster. GIScience and Remote Sensing, 58(7), 1130–1150. https://doi.org/10.1080/15481603.2021.1972712
Zhang, W., Qi, J., Wan, P., Wang, H., Xie, D., Wang, X., & Yan, G. (2016). An easy-to-use airborne LiDAR data filtering method based on cloth simulation. Remote Sensing, 8(6). https://doi.org/10.3390/rs8060501
Install
=======
*dendromatics* is available on `PyPI <https://pypi.org/project/dendromatics/>`_ and the full documentation can be consulted on `ReadTheDocs.io <https://dendromatics.readthedocs.io/en/latest/>`_
.. code-block:: console
python -m pip install dendromatics
The list of dependencies is available in the *pyproject.toml* file.
*dendromatics* relies on `hatch <https://github.com/pypa/hatch>`_
Depending on your version of Python and your OS, you might also need a C/C++ compiler to compile some of the mandatory dependencies (CSF and jakteristics).
But in any case you would not have to run the compiler by yourself, the build system will manage dependencies and compilation for you.
.. code-block:: console
python -m pip install hatch
You can run tests to ensure it works on your computer.
.. code-block:: console
hatch run cov
It is also possible to build doc locally.
.. code-block:: console
hatch run docs:build
hatch run docs:serve
and then go to `http://localhost:8000 <http://localhost:8000>`_ to browse it.
Raw data
{
"_id": null,
"home_page": "",
"name": "dendromatics",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.8",
"maintainer_email": "",
"keywords": "forestry,lidar,tls",
"author": "",
"author_email": "Carlos Cabo <carloscabo@uniovi.es>, Diego Laino <diegolainor@gmail.com>",
"download_url": "https://files.pythonhosted.org/packages/26/56/e439191a2c8ba37f7e92788a6b1ff055ec3154f5b45b1056ad57d2bbfe16/dendromatics-0.4.2.tar.gz",
"platform": null,
"description": ".. image:: https://github.com/3DFin/dendromatics/blob/main/docs/_static/dendromatics_logo.png\n :width: 300\n :align: center\n\nDendromatics\n============\n\n|docs| |pypi| |tests|\n\n.. |docs| image:: https://readthedocs.org/projects/dendromatics/badge/?version=latest\n :alt: Documentation Status\n :scale: 100%\n :target: https://dendromatics.readthedocs.io/en/latest/?badge=latest\n\n.. |pypi| image:: https://badge.fury.io/py/dendromatics.svg\n :alt: PyPI status\n :scale: 100%\n :target: https://badge.fury.io/py/dendromatics\n\n.. |tests| image:: https://github.com/3DFin/dendromatics/actions/workflows/test.yml/badge.svg\n :alt: Tests status\n :scale: 100%\n :target: https://github.com/3DFin/dendromatics/actions\n\n\nDescription\n===========\n\nThe *src* folder contains functionalities to detect the trees present in a terrestrial 3D point cloud from a forest plot, and compute individual tree parameters: tree height, tree location, diameters along the stem (including DBH), and stem axis. These are based on an updated version of the algorithm proposed by (Cabo et al., 2018).\n\nThe functionalities may be divided in four main steps:\n\n0. Height-normalization of the point cloud. \n1. Identification of stems among user-provided stripe.\n2. Tree individualization based on point-to-stems distances.\n3. Robust computation of stems diameter at different section heights.\n\nAlthough individual, somewhat independent functions are provided, they are designed to be used in a script that calls one after the other following the algorithm structure. An example script can be found in `example` folder.\n\n\nExamples\n========\n\n\nHeight-normalization\n--------------------\n\nAlmost all functions in the module expect a height-normalized point cloud to work as intended. If your point cloud is not height-normalized, you can do it in a simple way using some of the module functions. I'ts based on 'Cloth simulation Filter' (Zhang et al., 2016).\n\n.. code-block:: python\n \n import laspy\n import numpy as np\n import dendromatics as dm\n\n # Reading the point cloud\n filename_las = 'example_data.las' # your .las file\n entr = laspy.read(filename_las)\n coords = np.vstack((entr.x, entr.y, entr.z)).transpose()\n \n # Normalizing the point cloud\n dtm = dm.generate_dtm(clean_points)\n z0_values = dm.normalize_heights(coords, dtm)\n\n # adding the normalized heights to the point cloud\n coords = np.append(coords, np.expand_dims(z0_values, axis = 1), 1) \n\nIf the point cloud is noisy, you might want to denoise it first before generating the DTM.\n\n.. code-block:: python\n\n clean_points = dm.clean_ground(coords)\n\n\nIdentifying stems from a stripe\n-------------------------------\n\nSimply provide a stripe (from a height-normalized point cloud) as follows to iteratively 'peel off' the stems.\n\n.. code-block:: python\n\n # Defining the stripe\n lower_limit = 0.5\n upper_limit = 2.5\n stripe = coords[(coords[:, 3] > lower_limit) & (coords[:, 3] < upper_limit), 0:4]\n\n stripe_stems = dm.verticality_clustering(stripe, n_iter = 2) \n\n\nIndividualizing trees\n---------------------\n\nOnce the stems have been identified in the stripe, they can be used to individualize the trees in the point cloud.\n\n.. code-block:: python\n \n assigned_cloud, tree_vector, tree_heights = dm.individualize_trees(coords, stripe_stems) \n\n\nComputing sections along the stems\n----------------------------------\n\ncompute_sections() can be used to compute sections along the stems of the individualized trees.\n\n.. code-block:: python\n\n # Preprocessing: reducing the point cloud size by keeping only the points that are closer than some radius (expected_R) to the tree axes \n # and those that are whithin the lowest section (min_h) and the uppest section (max_h) to be computed.\n expected_R = 0.5\n min_h = 0.5 \n max_h = 25\n section_width = 0.02\n xyz0_coords = assigned_cloud[(assigned_cloud[:, 5] < expected_R) & (assigned_cloud[:, 3] > min_h) & (assigned_cloud[:,3] < max_h + section_width), :]\n \n stems = dm.verticality_clustering(xyz0_coords, n_iter = 2)[:, 0:6]\n \n # Computing the sections\n section_len = 0.2\n sections = np.arange(min_h, max_h, section_len) # Range of uniformly spaced values within the specified interval \n X_c, Y_c, R, check_circle, second_time, sector_perct, n_points_in = dm.compute_sections(stems, sections)\n\n\nTilt detection \n--------------\n\ntilt_detection() computes an 'outlier probability' for each section based on its tilting relative to neighbour sections and relative to the tree axis.\n\n.. code-block:: python\n \n outlier_prob = dm.tilt_detection(X_c, Y_c, R, sections)\n\n\nFor further examples and more thorough explanations, please check *example.py* script in */examples* folder.\n\n\nReferences\n==========\n\nCabo, C., Ord\u00f3\u00f1ez, C., L\u00f3pez-S\u00e1nchez, C. A., & Armesto, J. (2018). Automatic dendrometry: Tree detection, tree height and diameter estimation using terrestrial laser scanning. International Journal of Applied Earth Observation and Geoinformation, 69, 164\u2013174. https://doi.org/10.1016/j.jag.2018.01.011\n\n\nPrendes, C., Cabo, C., Ordo\u00f1ez, C., Majada, J., & Canga, E. (2021). An algorithm for the automatic parametrization of wood volume equations from Terrestrial Laser Scanning point clouds: application in Pinus pinaster. GIScience and Remote Sensing, 58(7), 1130\u20131150. https://doi.org/10.1080/15481603.2021.1972712 \n\n\nZhang, W., Qi, J., Wan, P., Wang, H., Xie, D., Wang, X., & Yan, G. (2016). An easy-to-use airborne LiDAR data filtering method based on cloth simulation. Remote Sensing, 8(6). https://doi.org/10.3390/rs8060501\n\n\nInstall\n=======\n\n*dendromatics* is available on `PyPI <https://pypi.org/project/dendromatics/>`_ and the full documentation can be consulted on `ReadTheDocs.io <https://dendromatics.readthedocs.io/en/latest/>`_\n\n.. code-block:: console\n \n python -m pip install dendromatics\n\nThe list of dependencies is available in the *pyproject.toml* file.\n\n*dendromatics* relies on `hatch <https://github.com/pypa/hatch>`_\n\nDepending on your version of Python and your OS, you might also need a C/C++ compiler to compile some of the mandatory dependencies (CSF and jakteristics). \nBut in any case you would not have to run the compiler by yourself, the build system will manage dependencies and compilation for you. \n\n.. code-block:: console\n \n python -m pip install hatch\n\nYou can run tests to ensure it works on your computer.\n\n.. code-block:: console\n \n hatch run cov\n\nIt is also possible to build doc locally.\n\n.. code-block:: console\n \n hatch run docs:build\n hatch run docs:serve\n\nand then go to `http://localhost:8000 <http://localhost:8000>`_ to browse it.\n",
"bugtrack_url": null,
"license": "",
"summary": "Automatic dendrometry and forest inventory for terrestrial point clouds",
"version": "0.4.2",
"project_urls": {
"Documentation": "https://dendromatics.readthedocs.io/en/latest/",
"Issues": "https://github.com/3DFin/dendromatics/issues",
"Source": "https://github.com/3DFin/dendromatics"
},
"split_keywords": [
"forestry",
"lidar",
"tls"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "dbce9f457ee76dc95c7bc0b8101a75621e509c37e325e850cdc2f2d58f21502e",
"md5": "3993db76d1ba49f20d55549546afe725",
"sha256": "b54ddca9e3a074a54c772090e1be6119b6a47883377cf531cf21a507c04f6dfd"
},
"downloads": -1,
"filename": "dendromatics-0.4.2-py3-none-any.whl",
"has_sig": false,
"md5_digest": "3993db76d1ba49f20d55549546afe725",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8",
"size": 42382,
"upload_time": "2024-03-15T07:59:24",
"upload_time_iso_8601": "2024-03-15T07:59:24.794037Z",
"url": "https://files.pythonhosted.org/packages/db/ce/9f457ee76dc95c7bc0b8101a75621e509c37e325e850cdc2f2d58f21502e/dendromatics-0.4.2-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "2656e439191a2c8ba37f7e92788a6b1ff055ec3154f5b45b1056ad57d2bbfe16",
"md5": "d5e00f90169c18d9250fe7c17ad737a5",
"sha256": "edb3308d40991ada96406737b9063cdf5cf016a4414bb57df219783f52d7a42d"
},
"downloads": -1,
"filename": "dendromatics-0.4.2.tar.gz",
"has_sig": false,
"md5_digest": "d5e00f90169c18d9250fe7c17ad737a5",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.8",
"size": 1056559,
"upload_time": "2024-03-15T07:59:27",
"upload_time_iso_8601": "2024-03-15T07:59:27.478936Z",
"url": "https://files.pythonhosted.org/packages/26/56/e439191a2c8ba37f7e92788a6b1ff055ec3154f5b45b1056ad57d2bbfe16/dendromatics-0.4.2.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-03-15 07:59:27",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "3DFin",
"github_project": "dendromatics",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "dendromatics"
}
