Skip to content

Mineral Detection toolbox in Python using Deep Learning based instance segmentation.

License

Notifications You must be signed in to change notification settings

norberttoth398/MinDet

Repository files navigation

MinDet

Documentation Status DOI Static Badge codecov

Repository for Deep Learning based petrography of igneous Plagioclase crystals based on circular polarised light images of thin sections. We make extensive use of the MMDetection library, with the work based on DetectoRS models.

Install

In order to install this private package you must be able to access it (which you can if you're reading this) and run have/create a python 3.7 environment for relevant package requirements (PyTorch can be a pain like that).

Ensure GCC is installed on your system.

Step 1

Create and activate environment:

eg conda create -n MinDetEnv python=3.7 

eg conda activate MinDetEnv

Step 2

Install required libraries (cluster nodes):

    wget https://download.pytorch.org/whl/cu110/torch-1.7.0%2Bcu110-cp37-cp37m-linux_x86_64.whl

pip install torch-1.7.0+cu110-cp37-cp37m-linux_x86_64.whl

pip install torchvision==0.8.0 torchaudio==0.7.0

    pip install openmim

    pip install mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu110/torch1.7.0/index.html

    mim install "mmdet<3.0.0"

Install required libraries (non-cluster):

pip install torch==1.7.0+cu110 torchvision==0.8.0 torchaudio==0.7.0 -f https://download.pytorch.org/whl/torch_stable.html

OR pip install torch==1.7.0 torchvision==0.8.0 torchaudio==0.7.0 (mac users)

pip install openmim

pip install mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu110/torch1.7.0/index.html

OR pip install mmcv-full -f https://download.openmmlab.com/mmcv/dist/cpu/torch1.7.0/index.html (mac users)

mim install "mmdet<3.0.0"

Install MinDet

Install using the following command:

pip install git+https://git@github.com/norberttoth398/MinDet

Supplementary Data

DOI

Cite

Please cite the following publication describing the present software:

Toth, N. and Maclennan, J. (2024) “MinDet1: A deep learning-enabled approach for plagioclase textural studies”, Volcanica, 7(1), pp. 135–151. doi: 10.30909/vol.07.01.135151.

About

Mineral Detection toolbox in Python using Deep Learning based instance segmentation.

Resources

License

Stars

Watchers

Forks

Packages

No packages published