Unfortunately, I cannot provide support for paddle any longer. Fortunately, there has been a lot of progress in the meantime, with regard to the usability of Mask R-CNN for custom applications, such as image-based particle detection. Therefore, I recommend to use the mmdetection framework, which has a large community, a thorough documentation and a huge model zoo. If you plan to train on your own data, this is a good place to get started.
This repository is a toolbox for the easy, deep learning-based primary particle size analysis of agglomerated, aggregated, partially sintered or simply occluded particles. It is the successor of the DeepParticleNet toolbox, which accompanies the following publication:
The utilized convolutional neural network is based on the Mask R-CNN architecture, developed by He et al.. It was implemented using PyTorch, torchvision and PyTorchLightning.
-
Install conda for your operating system.
-
Open a command line.
-
Clone this repository:
git clone https://github.com/maxfrei750/paddle.git -
Change into the folder of the repository:
cd paddle -
Create a new conda environment:
conda env create --file environment.yaml -
Activate the conda environment:
activate paddle -
Install paddle into the environment:
pip install --editable .
|
Tip
|
Creating the environment with conda can take a while. Using mamba can speed up the creation significantly.
|
-
Open a command line.
-
Activate the conda environment:
activate paddle -
Start JupyterLab:
jupyter laband click one of the links to access the JupyterLab server. -
In JupyterLab, navigate to the
paddle/demosfolder and choose a demo of your liking.
If you use this repository for a publication, then please cite it using the following bibtex-entry:
@article{Frei.2019,
author = {Frei, Max and Kruis, Frank Einar},
year = {2019},
title = {Image-Based Size Analysis of Agglomerated and Partially Sintered Particles via Convolutional Neural Networks},
url = {https://doi.org/10.1016/j.powtec.2019.10.020}
}