{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/reconstruction-of-three-dimensional-porous","title":"Reconstruction of three-dimensional porous media using generative adversarial neural networks","arxiv_id":"1704.03225","date":"2017-04-11","proceeding":null,"authors":["Lukas Mosser","Olivier Dubrule","Martin J. Blunt"],"abstract":"To evaluate the variability of multi-phase flow properties of porous media at\nthe pore scale, it is necessary to acquire a number of representative samples\nof the void-solid structure. While modern x-ray computer tomography has made it\npossible to extract three-dimensional images of the pore space, assessment of\nthe variability in the inherent material properties is often experimentally not\nfeasible. We present a novel method to reconstruct the solid-void structure of\nporous media by applying a generative neural network that allows an implicit\ndescription of the probability distribution represented by three-dimensional\nimage datasets. We show, by using an adversarial learning approach for neural\nnetworks, that this method of unsupervised learning is able to generate\nrepresentative samples of porous media that honor their statistics. We\nsuccessfully compare measures of pore morphology, such as the Euler\ncharacteristic, two-point statistics and directional single-phase permeability\nof synthetic realizations with the calculated properties of a bead pack, Berea\nsandstone, and Ketton limestone. Results show that GANs can be used to\nreconstruct high-resolution three-dimensional images of porous media at\ndifferent scales that are representative of the morphology of the images used\nto train the neural network. The fully convolutional nature of the trained\nneural network allows the generation of large samples while maintaining\ncomputational efficiency. Compared to classical stochastic methods of image\nreconstruction, the implicit representation of the learned data distribution\ncan be stored and reused to generate multiple realizations of the pore\nstructure very rapidly.","url_abs":"http://arxiv.org/abs/1704.03225v1","url_pdf":"http://arxiv.org/pdf/1704.03225v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"reconstruction-of-three-dimensional-porous","repo_url":"https://github.com/LukasMosser/PorousMediaGan","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"reconstruction-of-three-dimensional-porous","repo_url":"https://github.com/miniminisu/dcgan-code-cu-foam-3D","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"computational-efficiency","task_name":"Computational Efficiency"},{"task_slug":"image-reconstruction","task_name":"Image Reconstruction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}