rievil/crackenpy_dataset

CrackenPy dataset for training the models

This is a dataset of crack images on alkali-activated materials created as part of a project funded by the Czech Science Foundation under grant number 22-02098S, titled “Experimental analysis of the shrinkage, creep, and cracking mechanism of materials based on alkali-activated slag.” The dataset consists of 1,207 photos and masks of test specimens at various stages of degradation due to shrinkage cracks. Each individual photo is accompanied by a grayscale mask that includes the following image classes:

• Background
• Matrix
• Crack
• Pore

Supplementary Data

The dataset was developed for training a convolutional neural network for image segmentation. The images in the dataset consist of sub-images with a resolution of 416x416 pixels. Original full-resolution images of the specimens are also available for download upon request from the authors.

Software

The dataset was used to develop the CrackenPy library in Python, which enables efficient application of the pre-trained model on any photos of test specimens. The dataset, library, and pre-trained model are publicly accessible on GitHub and Hugging Face.

• Library
• Model
• Dataset

The library can be installed using the following command:

    pip install crackenpy

The dataset itself can be downloaded from The Hugging face web upon filling the form.

Citation

If you use CrackenPy tools (library, model, dataset), please cite our work:

APA style: Dvorak, R., Bilek, V., Krc, R., & Kucharczykova, B. (2024). CrackenPy: Image segmentation tool for semantic segmentation of building material surfaces using deep learning [Computer software]. https://doi.org/10.5281/zenodo.13969747

@misc {richard_dvorak_2024,
    author       = { {Richard Dvorak} },
    title        = { crackenpy (Revision 04ed02c) },
    year         = 2024,
    url          = { https://huggingface.co/rievil/crackenpy },
    doi          = { 10.57967/hf/3295 },
    publisher    = { Hugging Face }
}

@software {Dvorak_CrackenPy_Image_segmentation_2024,
    author       = {Dvorak, Richard and Bilek, Vlastimil and Krc, Rostislav and Kucharczykova, Barbara},
    doi          = {10.5281/zenodo.13969747},
    month        = oct,
    title        = {{CrackenPy: Image segmentation tool for semantic segmentation of building material surfaces using deep learning}},
    url          = {https://github.com/Rievil/CrackenPy},
    year         = {2024}
}

@misc {richard_dvorak_2024,
    author       = { {Richard Dvorak} },
    title        = { crackenpy_dataset (Revision ce5c857) },
    year         = 2024,
    url          = { https://huggingface.co/datasets/rievil/crackenpy_dataset },
    doi          = { 10.57967/hf/3496 },
    publisher    = { Hugging Face }
}

License

The CrackenPy Dataset is made freely available to academic and non-academic entities for non-commercial purposes, such as academic research, teaching, scientific publications, or personal experimentation. Permission is granted to use the data provided that you agree to the following terms:

1. The dataset is provided “AS IS,” without any express or implied warranty. Although every effort has been made to ensure accuracy, we (Brno University of Technology) do not accept any responsibility for errors or omissions.

2. Any work that makes use of the dataset must include a reference to the CrackenPy Dataset. For research papers or other media, please link to the CrackenPy Dataset webpage.

3. You may not distribute this dataset or modified versions of it. However, it is permissible to distribute derivative works, as long as they are abstract representations of this dataset (such as models trained on it or additional annotations that do not directly include any of our data) and do not allow recovery of the dataset or similar data.

4. You may not use the dataset or any derivative work for commercial purposes, including licensing, selling, or using the data for commercial gain.

5. All rights not expressly granted to you are reserved by us (Brno University of Technology).

Acknowledgement

We would like to thank the staff and students of the Faculty of Civil Engineering at Brno University of Technology for preparing the dataset, labeling the images and providing space for the research.