Fracture network segmentation (doi:10.18419/darus-1847)

View:

Part 1: Document Description
Part 2: Study Description
Part 5: Other Study-Related Materials
Entire Codebook

Fracture network segmentation

doi:10.18419/darus-1847

DaRUS

2021-07-07

1

Lee, Dongwon; Nikolaos, Karadimitriou; Steeb, Holger, 2021, "Fracture network segmentation", https://doi.org/10.18419/darus-1847, DaRUS, V1

Fracture network segmentation

doi:10.18419/darus-1847

Lee, Dongwon (University of Stuttgart)

Nikolaos, Karadimitriou (University of Stuttgart)

Steeb, Holger (University of Stuttgart)

University of Stuttgart

2021

327154368

DaRUS

University of Stuttgart

Steeb, Holger

Lee, Dongwon

2021-05-11

2021-05-26

https://doi.org/10.18419/darus-1847

Computer and Information Science, Earth and Environmental Sciences, Engineering, Micro fractures, Image segmentation, Machine learning, Thermally treated Carrara marble

Software

This dataset contains the codes to reproduce the five different segmentation results of the paper Lee et al (2021). The original dataset before applying these segmentation codes could be found in Ruf & Steeb (2020). The adopted segmentation methods in order to identify the micro fractures within the original dataset are the Local threshold, Sato, Chan-Vese, Random forest and U-net model. The Local threshold, Sato and U-net models are written in Python. The codes require a version above Python 3.7.7 with tensorflow, keras, pandas, scipy, scikit and numpy libraries. The workflow of the Chan-Vese method is interpreted in Matlab2018b. The result of the Random forest method could be reproduced with the uploaded trained model in an open source program ImageJ and trainableWeka library. For further details of operation, please refer to the readme.txt file.

program source code

Ruf, M., & Steeb, H. (2020). micro-XRCT data set of Carrara marble with artificially created crack network: fast cooling down from 600°C. DaRUS. <a href="https://doi.org/10.18419/DARUS-682">https://doi.org/10.18419/DARUS-682</a>

Lee, D., Karadimitriou, N., Ruf, M., & Steeb, H. (2021). Detecting micro fractures with X-ray computed tomography.

2103.12821

Lee, D., Karadimitriou, N., Ruf, M., & Steeb, H. (2021). Detecting micro fractures with X-ray computed tomography.

Label:

Chan_vese.m

Text:

This file contains the workflow of Chan-Vese method

text/plain

Label:

classifier.model

Text:

This is the trained model with the Random forest method. This file could be loaded in the open source program ImageJ with TrainableWeka library. The predictions can be generated with processing.bsh code.

application/octet-stream

Label:

Frac_Unet_postprocess.py

Text:

With the code, the predictions of trained model for given input files can be generated and saved.

text/plain

Label:

Frac_Unet_preprocess.py

Text:

Preparation of training data can be done with the code

text/plain

Label:

Frac_Unet_training.py

Text:

With the pre-processed training data, the code performs training of U-net model and saves a trained model after training.

text/plain

Label:

img_seg.py

Text:

This file contains the load/save functions and core functions of the Sato and the Local threshold methods.

text/plain

Label:

Local_threshold.py

Text:

This file contains the workflow of the local threshold method

text/plain

Label:

post_random_forest.py

Text:

The predictions of the Random forest method can be merged into the original size of image with the code.

text/plain

Label:

processing.bsh

Text:

Prediction generator of the random forest method. The input file path, save path and trained model have to be chosen.

application/octet-stream

Label:

readme.txt

Text:

This file contains instruction of how to operate the codes.

text/plain

Label:

Sato.py

Text:

This file contains the workflow of the Sato method

text/plain

Label:

trained_model_U_net_27.hdf5

Text:

Trained model of the U-net model. This file is required to reproduce the results.

application/x-hdf5

Label:

unet_v2_mod.py

Text:

The code to create the U-net architecture

text/plain