AlignmentTool

The AlignmentTool was developed to offer a highly supervised and interactive method to perform cadastres' alignment. But it could be used for other image alignment tasks!

The backend processes are mainly handled through OpenCv, the data structure is managed with NetworkX, and the Notebook interface relies on jupyter-innotater. It is provided as a Juyter Notebook entitled Tool.ipynb.

Find out more about this project on this Wiki.

Set up

Run the following command in terminal to install the recquired packages:

pip install -r requirements.txt

Functionalities

The Tool.ipynb notebook allows to:

• preprocess the images
  • rename
  • rotate
• perform template matching on pairs of images
  • select the template and the matching area directly on the images
  • evaluate the results and give feedback to the machine
• compose large images based on pairwise matches

The notebook is documented to guide the user and ease its use.

See on the gif below what the process can look like:

Structure

.
|
├── Tool.ipynb                            # Main Notebook
|
├── cadastre_matching.py                  # Underlying data processing module
|
├── innotation_functions.py               # Module to handle jupyter-innotater APIs calls
|
├── demo_material/                        # Images and gif
│   └── ...                                 # 
├── README.md
└── requirements.txt

Team

• Marin — @MazarinDH (marin.piguet@epfl.ch)
• Noé — @noe-d (noe.durandard@epfl.ch)