Recursive Water and Land Mapping

This code has been implemented in Python 3.6.13, using the IDE Pycharm 2021.2.3.

Project Structure

The project is structured as follows.

• README.md

• requirements.txt

• ./benchmark/

  • ./deepwatermap_main/ contains part of the deepwatermap algorithm open source code shared in this GitHub repository. An __init__.py file has been added to this directory to treat it as a module.

  • ./watnet/ contains part of the WatNet algorithm open source code shared in this GitHub repository. An __init__.py file has been added to this directory to treat it as a module.

  • benchmark.py includes a copy of two functions from ./deepwatermap_main/inference.py. The main function contains one change.

• ./logs/ contains the log files generated at every code execution.

• ./tools/ contains useful functions.

  • operations.py

  • path_operations.py

  • spectral_index.py

• configuration.py contains the classes Debug and Config. Some configuration settings must be changed when executing the code with data that is different to the one provided by the authors.

• main.py contains the main logic and flow of the code.

• bayesian_recursive.py contains the core of the recursive bayesian algorithm for classification.

• figures.py includes functions to plot the generated results.

• image_reader.py contains the abstract class ImageReader and the class ReadSentinel2 which allows the user to read images from a dataset of Sentinel2 images, such as the one provided by the authors of this code.

• ./training_data/ contains pickle files with saved data from the training stage. If wanting to train the models from scratch, it should be indicated in the Debug class from configuration.py. Data has been stored in this file because the training stage takes long.

• training.py contains functions that are linked to the training stage.

• evaluation.py contains functions that are linked to the evaluation stage.

Installation and Running

Required packages can be installed using the Python package installer pip:

pip install -r requirements.txt

Download the data from our Dataset by...

Generate the results presented in our publication by... calling the main function or for more detailed guidance check the Jupyter Notebook file...

References

We would like to highlight that the open source codes of the algorithms used for benchmarking purposes were provided by their respective authors:

• The DeepWaterMap algorithm presented in this GitHub repository, by L. F. Isikdogan, A.C. Bovik and P. Passalacqua. This algorithm for Water Mapping is presented in the following publications:

  • Seeing Through the Clouds With DeepWaterMap

  • Surface Water Mapping by Deep Learning

• The WatNet algorithm presented in this GitHub repository, by Xin Luo, Xiaohua Tong and Zhongwen Hu. This algorithm for Water Mapping is presented in the publication An applicable and automatic method for earth surface water mapping based on multispectral images.

Authors

• Bhavya Duvvuri, from The Beighley Lab (Sustainable Water Resources | Resilient Wet Infrastructure) at Northeastern University, Boston (MA).
• Helena Calatrava, from the Signal Processing, Imaging, Reasoning and Learning (SPIRAL) Group also at Northeastern University, Boston (MA).