Installation

Follow up this tutorial to set up the landslide detection tool

🌄 Intro

Landslide_detector is built upon Python Based Open Source System for Geographic Object Based Image Analysis ([GEOBIA])(https://www.sciencedirect.com/science/article/pii/S0098300413002288) that uses the following libraries: GDAL, RSGISLib, TuiView, RIOS.

Minimum requirements (Win)

• 💻 Windows OS x64
• 🐍 Python Anaconda
• 📦 Python (>= 3.5)
• 📦 NumPy (>= 1.11.0)
• 📦 SciPy (>= 0.17.0)
• 📦 joblib (>= 0.11)

Installation

    conda create --name ld --file Requirements.txt

⚠️ If this does not work, follow next steps to manually install GEOBIA, Scikit learn and additional dependencies.

Alternative Installation

1. Install Python Anaconda

2. Within the conda terminal, install gdal running:

   conda install -c conda-forge gdal 

For more information visit https://anaconda.org/conda-forge/gdal

3. Install RSGISLib using:

  conda create -n rsgislib -c conda-forge rsgislib   

  activate rsgislib

4. TuiView: for viewing and manipulating RATs (Raster Attribute Tables)

• To install this package with conda run:

conda config --add channels conda-forge

conda create -n myenv tuiview

conda activate myenv

For more information visit http://tuiview.org/

• To open the raster and/or segmented images files:

Within the conda terminal (in myenv) run: tuiview

5. Raster I/O Simplification (RIOS): for reading, writing and classifying attributed objects

To install this package with conda run:

conda install -c conda-forge rios 

For more information visit https://anaconda.org/conda-forge/rios

6. Scikit Learn installation

For more information visit https://scikit-learn.org/stable/install.html

To install this package with conda run:

conda install scikit-learn

7. Install pyyaml:

conda install -c anaconda pyyaml