{"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/learning-to-reconstruct-crack-profiles-for","title":"Learning to Reconstruct Crack Profiles for Eddy Current Nondestructive Testing","arxiv_id":"1910.08721","date":"2019-10-28","proceeding":null,"authors":[],"abstract":"Eddy current testing (ECT) is one of the most popular Nondestructive Testing\n(NDT) techniques, especially for conductive materials. Reconstructing the crack\nprofile from measured EC signals is one of the main goals of ECT. This task is\nhighly challenging, as the EC signals are nonlinear responses resulted from the\npresence of cracks, and reconstructing the crack profile requires establishing\nthe forward model of the nonlinear electromagnetic dynamics and solving its\ninverse problem, which is an ill-posed numerical optimization problem. Instead\nof solving the inverse problem numerically, we propose to directly learn the\ninverse mapping from EC signals to crack profiles with a deep encoder-decoder\nconvolutional neural network named EddyNet. EddyNet is trained on a set of\nrandomly generated crack profiles and the corresponding simulated EC responses\ngenerated from a realistic forward model. On the held-out test data, EddyNet\nachieved a mean absolute error of 0.198 between predicted profiles and ground\ntruth ones. Qualitatively, the geometries of predicted profiles are visually\nsimilar to the ground truth profiles. Our method greatly reduces the usual\nreliance on domain experts, and the reconstruction is extremely fast both on\nGPUs and on CPUs. The source code of EddyNet is released on\nhttps://github.com/askerlee/EddyNet.","url_abs":"http://arxiv.org/abs/1910.08721v2","url_pdf":"http://arxiv.org/pdf/1910.08721v2.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":"learning-to-reconstruct-crack-profiles-for","repo_url":"https://github.com/askerlee/EddyNet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}