Superpixel segmentations for thin sections: Evaluation of methods to enable the generation of machine learning training data sets

Computers & Geosciences, vol. 170, pp. 105232

Abstract

Training data is the backbone of developing either Machine Learning (ML) models or specific Deep Learning (DL) algorithms. The lack of well-labeled training image data has significantly impeded the development of novel DL methods like Convolutional Neural Networks (CNNs) in mineral thin section images identification. However, image annotation, especially pixel-wise annotation is always a costly process. Manually creating dense semantic labels for rock thin section images has been long considered as an unprecedented challenge in view of the ubiquitous variety and complexity of minerals in thin sections. To speed up the annotation, we propose a human–computer collaborative pipeline in which superpixel segmentation is used as a boundary extractor to avoid hand delineation of instances boundaries. The pipeline consists of two steps: superpixel segmentation using MultiSLIC, and superpixel labeling through a specific-designed tool. We use a cutting-edge methodology Virtual Petroscopy (ViP) for automatic image acquisition. Bentheimer sandstone sample is used to conduct performance testing of the pipeline. Three standard error metrics are used to evaluate the performance of MultiSLIC. The result indicates that MultiSLIC is able to extract compact superpixels with satisfying boundary adherence given multiple input images. According to our test results, large and complex thin section images with pixel-wise accurate labels can be annotated with the labeling tool more efficiently than in a conventional, purely manual work, and generate data of high quality.

Authors 7

1. Jiaxin Yu Aachen Department of Computational Geoscience

   RWTH Aachen University

   Affiliation as printed

   RWTH-Aachen, Geoscience, Department of Computational Geoscience, Geothermics and Reservoir Geophysics, Mathieustrasse 30, 52074 Aachen, Germany

2. Florian Wellmann corresponding Aachen Department of Computational Geoscience Terranigma Solutions GmbH

   Fraunhofer Research Institution for Energy Infrastructures and Geotechnologies IEG · RWTH Aachen University

   Affiliation as printed

   Fraunhofer IEG – Fraunhofer Research Institution for Energy Infrastructures and Geothermal Systems, Am Hochschulcampus 1/IEG, 44801 Bochum, Germany

   RWTH-Aachen, Geoscience, Department of Computational Geoscience, Geothermics and Reservoir Geophysics, Mathieustrasse 30, 52074 Aachen, Germany

   Terranigma Solutions GmbH, Laurentiusstrasse 59, 52072 Aachen, Germany

3. Simon Virgo Aachen Department of Computational Geoscience Terranigma Solutions GmbH

   RWTH Aachen University

   Affiliation as printed

   RWTH-Aachen, Geoscience, Department of Computational Geoscience, Geothermics and Reservoir Geophysics, Mathieustrasse 30, 52074 Aachen, Germany

   Terranigma Solutions GmbH, Laurentiusstrasse 59, 52072 Aachen, Germany

4. Marven von Domarus Aachen Visual Computing Institute Computer Vision Group

   RWTH Aachen University

   Affiliation as printed

   RWTH-Aachen ,Visual Computing Institute, Computer Vision group, Mies-van-der-Rohe Strasse 15, 52074 Aachen, Germany

5. Mingze Jiang Aachen Map – Microstructure and Pores GmbH Department of Computational Geoscience

   Microstructure and Pores · RWTH Aachen University

   Affiliation as printed

   MaP Microstructure and Pores GmbH, Junkerstraße 93, 52064 Aachen, Germany, Germany

   RWTH-Aachen, Geoscience, Department of Computational Geoscience, Geothermics and Reservoir Geophysics, Mathieustrasse 30, 52074 Aachen, Germany

6. Joyce Schmatz Aachen Map – Microstructure and Pores GmbH

   Microstructure and Pores · RWTH Aachen University

   Affiliation as printed

   MaP Microstructure and Pores GmbH, Junkerstraße 93, 52064 Aachen, Germany, Germany

   RWTH-Aachen, Lehr- und Forschungsgebiet für Geologie-Endogene Dynamik, Germany

7. Bastian Leibe Aachen Visual Computing Institute Computer Vision Group

   RWTH Aachen University

   Affiliation as printed

   RWTH-Aachen ,Visual Computing Institute, Computer Vision group, Mies-van-der-Rohe Strasse 15, 52074 Aachen, Germany