Point cloud pathdinfing tool

About The Project

This is a python tool that uses data clustering techniques to identify relevant locations inside Lidar point cloud data. The output is a JSON file containing targets in the shape of coordinates and positions that the camera must go in order to have a clear view over the targets.

It was developed as part of my Bachelor Thesis work for my HES-SO B.S. in Computer Science & Engineering, oriented in Software Development in HEIG-VD, Switzerland.

This tool can be useful in application with this adaptation of the Potree tool that allows the import of such JSON files and the visualization of camera paths.

Built With

This section should list any major frameworks/libraries used to bootstrap your project. Leave any add-ons/plugins for the acknowledgements section. Here are a few examples.

• 
• 
• 
• 
• 

Getting Started

This project is a very simple python file that can be executed to generate a JSON with the targets and positions.

Prerequisites

To be sure everything is going to work properly, be sure you have Python version >= 3.8

• Get python version

  python --version

Installation

Then you must install the required libraries to be able to use the script properly.

1. Clone the repo

   git clone https://github.com/MazieroMarco/bt-pathfinding-tool.git

2. Install python packages

   pip install -r requirements.txt

3. Run the pathfinder script

   python pathfinder.py my_file.las -o ./my_output.json

Usage

• Run simple clustering with default params (5 targets, 0.1% data sample size)

  python pathfinder.py my_file.las

• Run simple clustering with custom params (10 targets, 0.4% data sample size, custom epsilon, custom output)

  python pathfinder.py my_file.las -o ./output.json -p 10 -q 0.4 -e 0.9

More details are avilable in the comand line tool help menu by using the -h option.

• Help menu

  python pathfinder.py -h
  usage: pathfinder.py [-h] [--output DIR] [--poi N] [--quantity N] [--epsilon N] INPUT
  
  Finds interesting locations and a camera path inside a given LAS point cloud data file.
  
  positional arguments:
    INPUT                 The path of the input LAS data file
  
  optional arguments:
    -h, --help            show this help message and exit
    --output DIR, -o DIR  The output directory for the generated JSON file
    --poi N, -p N         The amount of points of interest to output
    --quantity N, -q N    The proportion of points to keep in the working data sample [0 < q < 1].
                          Warning, a big number slows down the algorithm.
    --epsilon N, -e N     The epsilon parameter used for the data clustering. This parameter is
                          approximated if no value is given.
  

Contributing

Contributions are what make the open source community such an amazing place to learn, inspire, and create. Any contributions you make are greatly appreciated.

License

Distributed under the MIT License. See LICENSE for more information.

Contact

Marco Maziero - maziero.marco@outlook.com