install.md

Installation

To use our code, first download this repository and initialize the submodules:

git clone https://github.com/GrumpyZhou/image-matching-toolbox.git

# Install submodules non-recursively
cd image-matching-toolbox/
git submodule update --init

Next, download the pretained models and place them to the correct places by running the followings:

cd pretrained/
bash download.sh

Setup Running Environment

Following the steps to setup the ready environment to run the matching toolbox. The code has been tested on Ubuntu 18.04 with Python 3.7 + Pytorch 1.7.0 + CUDA 10.2.

1. Create the immatch virtual environment

conda env create -f environment.yml
conda activate immatch

Notice, the immatch conda env allows to run all supported methods expect for SparseNCNet. In order to use it, please install its required dependencies according to its official installation,

2. Install the immatch toolbox as a python package

# Install immatch toolbox
cd image-matching-toolbox/
python setup.py develop

The developing mode allows you to change the code without re-installing it in the environment. You can also install the matching toolbox to any environment to use it for your other projects. To uninstall it from an environment:

pip uninstall immatch

3. Install pycolmap

This package is essential for evaluations on localization benchmarks.

# Install  pycolmap 
pip install git+https://github.com/mihaidusmanu/pycolmap

In case https link doesnt work, you can install it directly for Python 3.7 and Python 3.8 via pypi:

pip install pycolmap

4. Update immatch environment when needed

Incase more packages are needed for new features, one can update your created immatch environment:

Option 1: add new libs into setup.py (Recommended & Faster)

# Update immatch toolbox
cd image-matching-toolbox/
python setup.py develop

Option 2: add new libs into environment.yml

conda activate immatch
conda env update --file environment.yml --prune