{"id":8398,"date":"2024-11-20T12:51:07","date_gmt":"2024-11-20T12:51:07","guid":{"rendered":"https:\/\/lcm.web-email.at\/measurement-of-hairline-cracks-using-deep-learning-2\/"},"modified":"2025-08-13T10:32:26","modified_gmt":"2025-08-13T10:32:26","slug":"measurement-of-hairline-cracks-using-deep-learning-2","status":"publish","type":"post","link":"https:\/\/lcm.web-email.at\/en\/measurement-of-hairline-cracks-using-deep-learning-2\/","title":{"rendered":"Measurement of hairline cracks using deep learning"},"content":{"rendered":"\n<section class=\"wp-block-group section-post has-background\" style=\"background-color:#f3f3f4;margin-top:0;margin-bottom:0;padding-top:var(--wp--preset--spacing--50);padding-bottom:var(--wp--preset--spacing--50)\"><div class=\"wp-block-group__inner-container is-layout-constrained wp-container-core-group-is-layout-5c411644 wp-block-group-is-layout-constrained\">\n<div style=\"height:50px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<h2 class=\"wp-block-heading\">Measurement of <mark>hairline cracks <\/mark>using <mark>deep learning<\/mark><\/h2>\n\n\n\n<div style=\"height:25px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns section-post-wrap is-layout-flex wp-container-core-columns-is-layout-28f84493 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\">\n<figure class=\"wp-block-image size-full is-style-default\"><img decoding=\"async\" width=\"2560\" height=\"1707\" src=\"https:\/\/lcm.web-email.at\/wp-content\/uploads\/2025\/01\/Post_Projekt_Haarisse_KI.png\" alt=\"Post Project Haarisse AI\" class=\"wp-image-1354\" srcset=\"https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/01\/Post_Projekt_Haarisse_KI.png 2560w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/01\/Post_Projekt_Haarisse_KI-300x200.png 300w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/01\/Post_Projekt_Haarisse_KI-1024x683.png 1024w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/01\/Post_Projekt_Haarisse_KI-768x512.png 768w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/01\/Post_Projekt_Haarisse_KI-1536x1024.png 1536w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/01\/Post_Projekt_Haarisse_KI-2048x1366.png 2048w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/01\/Post_Projekt_Haarisse_KI-178x119.png 178w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/01\/Post_Projekt_Haarisse_KI-124x83.png 124w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/01\/Post_Projekt_Haarisse_KI-82x55.png 82w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/01\/Post_Projekt_Haarisse_KI-24x16.png 24w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/01\/Post_Projekt_Haarisse_KI-182x121.png 182w\" sizes=\"(max-width: 2560px) 100vw, 2560px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\">\n<h3 class=\"wp-block-heading\">Image-based quality assurance<\/h3>\n\n\n\n<p>Cracks are to be detected and measured on steel strips with a large data volume of 400 microscope images per sample. The cracks in the image are very thin, approximately 1 pixel wide. In addition, artifacts make crack detection difficult, which is why conventional image processing methods do not offer an exact and robust solution.  <\/p>\n\n\n\n<p>Sabrina Fleischanderl et al, \u201cCNN-based crack detection in oxide layers of hot rolled steel sheet samples for the validation of a pickling process model\u201d, Proc. 3rd Symp. on Pattern Recognition and Applications, 2022.<\/p>\n<\/div>\n<\/div>\n\n\n\n<div style=\"height:50px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div><\/section>\n\n<div style=\"height:25px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n    <section  class=\"section-desc\">\n\t\t<div class=\"container\">\n            <div class=\"section-desc-wrap\">\n\t\t\t\t\t\t\t\t\t<div class=\"col\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t<h2> \n\t\t\t\t\t\t\t\tProject <mark>requirements <\/mark>\t\t\t\t\t\t\t<\/h2>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<div class=\"block-desc\">\n\t\t\t\t\t\t\t\t<ul>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<li>Cracks should not only be detected, but also measured (density, vertical distance in the image)<\/li>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<li>Large amount of data (400 microscope images per sample)<\/li>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<li>The cracks are very thin in the image (width ~ 1 px), and other artifacts make crack detection more difficult<\/li>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<li>Classic image processing cannot provide an exact and robust solution<\/li>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/ul>\n\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t\t<div class=\"col\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t<h2> \n\t\t\t\t\t\t\t\tThe <mark>solution<\/mark>\t\t\t\t\t\t\t<\/h2>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<div class=\"block-desc\">\n\t\t\t\t\t\t\t\t<ul>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<li>Precise and stable segmentation of cracks with neural network (UNET)<\/li>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<li>60 images were annotated by hand<\/li>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<li>Wireless connection of each measuring point to the relay station<\/li>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<li>In addition, a freely available annotated data set (ISBI 2012 Challenge, cell membrane data) was used for training<\/li>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<li>Precise solution with minimal effort<\/li>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/ul>\n\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t            <\/div>\n        <\/div>\n    <\/section>\n\n<div style=\"height:25px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n    <div  class=\"section-card-person bg-gray-300-30\">\n        \t\t<div class=\"container\">\n\t\t\t<div class=\"card-person\">\n                <div class=\"card-person-img\">\n                                            <img decoding=\"async\" width=\"222\" height=\"222\" src=\"https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/02\/Ansprechpartner_Schimbaeck.png\" class=\"attachment-thumbnail_229x229 size-thumbnail_229x229\" alt=\"\" srcset=\"https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/02\/Ansprechpartner_Schimbaeck.png 222w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/02\/Ansprechpartner_Schimbaeck-150x150.png 150w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/02\/Ansprechpartner_Schimbaeck-50x50.png 50w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/02\/Ansprechpartner_Schimbaeck-119x119.png 119w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/02\/Ansprechpartner_Schimbaeck-83x83.png 83w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/02\/Ansprechpartner_Schimbaeck-55x55.png 55w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/02\/Ansprechpartner_Schimbaeck-16x16.png 16w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/02\/Ansprechpartner_Schimbaeck-182x182.png 182w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/02\/Ansprechpartner_Schimbaeck-90x90.png 90w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/02\/Ansprechpartner_Schimbaeck-120x120.png 120w\" sizes=\"(max-width: 222px) 100vw, 222px\" \/>                                                                <div class=\"logo\"><img decoding=\"async\" width=\"35\" height=\"15\" src=\"https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2024\/11\/logo-short.svg\" class=\"attachment-thumbnail_35x16 size-thumbnail_35x16\" alt=\"\" \/><\/div>\n                                    <\/div>\n\t\t\t\t\t\t\t\t\t<div class=\"card-person-body\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t<span class=\"note\">Contact Person<\/span>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<h2 class=\"h3\">Erwin Schimb\u00e4ck<\/h2>\n\t\t\t\t\t\t                                                    <span class=\"position\">Business Area Manager Sensors &amp; Communication<\/span>\n                                                                            <div class=\"contact-hold\">\n                                                                    <div class=\"frame\">\n                                                                                                                                    <div class=\"text-hold\">\n                                                    <a href=\"tel:4373224686138\">\n                                                        <span class=\"icon\">\n                                                            <svg xmlns=\"http:\/\/www.w3.org\/2000\/svg\" width=\"32\" height=\"32\" viewBox=\"0 0 32 32\"><path d=\"M22 20c-2 2-2 4-4 4s-4-2-6-4-4-4-4-6 2-2 4-4-4-8-6-8-6 6-6 6c0 4 4.109 12.109 8 16s12 8 16 8c0 0 6-4 6-6s-6-8-8-6z\"><\/path><\/svg>\n                                                        <\/span>\n                                                        <span class=\"text\">+43 732 2468 6138<\/span>\n                                                    <\/a>\n                                                <\/div>\n                                                                                                                                        <div class=\"text-hold\">\n                                                    <a href=\"mailto:erwi&#110;&#46;&#115;ch&#105;m&#98;a&#101;&#99;k&#64;&#108;c&#109;.&#97;t\">\n                                                        <span class=\"icon\">\n                                                            <svg xmlns=\"http:\/\/www.w3.org\/2000\/svg\" width=\"28\" height=\"28\" viewBox=\"0 0 28 28\"><path d=\"M28 11.094v12.406c0 1.375-1.125 2.5-2.5 2.5h-23c-1.375 0-2.5-1.125-2.5-2.5v-12.406c0.469 0.516 1 0.969 1.578 1.359 2.594 1.766 5.219 3.531 7.766 5.391 1.313 0.969 2.938 2.156 4.641 2.156h0.031c1.703 0 3.328-1.188 4.641-2.156 2.547-1.844 5.172-3.625 7.781-5.391 0.562-0.391 1.094-0.844 1.563-1.359zM28 6.5c0 1.75-1.297 3.328-2.672 4.281-2.438 1.687-4.891 3.375-7.313 5.078-1.016 0.703-2.734 2.141-4 2.141h-0.031c-1.266 0-2.984-1.437-4-2.141-2.422-1.703-4.875-3.391-7.297-5.078-1.109-0.75-2.688-2.516-2.688-3.938 0-1.531 0.828-2.844 2.5-2.844h23c1.359 0 2.5 1.125 2.5 2.5z\"><\/path><\/svg>\n                                                        <\/span>\n                                                        <span class=\"text\">erwin.schimbaeck@lcm.at<\/span>\n                                                    <\/a>\n                                                <\/div>\n                                                                                                                        <\/div>\n                                                                                                    <a class=\"btn\" href=\"mailto:&#101;r&#119;i&#110;&#46;&#115;chimb&#97;&#101;&#99;k&#64;&#108;c&#109;&#46;at\">Send request<\/a>\n                                                            <\/div>\n                        \t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n            <div class=\"section-decor\"><img decoding=\"async\" src=\"https:\/\/lcm.web-email.at\/wp-content\/themes\/linzcenterofmech\/assets\/images\/img-21.png\" alt=\"image description\" width=\"712\" height=\"414\" role=\"img\"><\/div>\n        <\/div>\n            <\/div>\n\t<section  class=\"section-main\">\n\t\t<div class=\"container\">\n\t\t\t\t\t\t\t<div class=\"section-title\">\n\t\t\t\t\t\t\t\t\t\t\t<h2>Further <mark>reference projects<\/mark><\/h2>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t\t\t<div class=\"card-wrap\">\n\t\t\t\t\t\t\t\t\t\t\t<div class=\"card ffg-success-story-en sensors-and-communication \">\n\t<div class=\"card-img\">\n\t\t<img decoding=\"async\" width=\"389\" height=\"240\" src=\"https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/09\/Fraese_Hackschnitzel-389x240.jpeg\" class=\"attachment-thumbnail_389x240 size-thumbnail_389x240 wp-post-image\" alt=\"\" srcset=\"https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/09\/Fraese_Hackschnitzel-389x240.jpeg 389w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/09\/Fraese_Hackschnitzel-24x16.jpeg 24w\" sizes=\"(max-width: 389px) 100vw, 389px\" \/>\t\t\t<\/div>\n\t<div class=\"card-body\">\n\t\t<div class=\"frame\">\n\t\t\t<h4><a href=\"https:\/\/lcm.web-email.at\/en\/multifunctional-milling-machine\/\" rel=\"bookmark\">Multifunctional milling machine<\/a><\/h4>\t\t\t<p>Discover how customized functional steels improve the properties of sheet steel and open up new application possibilities.<\/p>\n\t\t<\/div>\n\t\t<a class=\"link-more\" href=\"https:\/\/lcm.web-email.at\/en\/multifunctional-milling-machine\/\">Read more<\/a>\n\t<\/div>\n<\/div>\t\t\t\t\t\t\t\t\t\t\t<div class=\"card sensors-and-communication modeling-and-simulation research-project sensor-systems-and-measurement-services vibration-technology \">\n\t<div class=\"card-img\">\n\t\t<img decoding=\"async\" width=\"389\" height=\"240\" src=\"https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/08\/Bild_Beitrag_Schwingungstilger-389x240.png\" class=\"attachment-thumbnail_389x240 size-thumbnail_389x240 wp-post-image\" alt=\"\" \/>\t\t\t<\/div>\n\t<div class=\"card-body\">\n\t\t<div class=\"frame\">\n\t\t\t<h4><a href=\"https:\/\/lcm.web-email.at\/en\/adjustable-passive-vibration-damper\/\" rel=\"bookmark\">Adjustable passive vibration damper<\/a><\/h4>\t\t\t<p>Adjustable passive vibration damper Effective vibration reduction for buildings Buildings are exposed to a variety of loads, for example from&#8230;<\/p>\n\t\t<\/div>\n\t\t<a class=\"link-more\" href=\"https:\/\/lcm.web-email.at\/en\/adjustable-passive-vibration-damper\/\">Read more<\/a>\n\t<\/div>\n<\/div>\t\t\t\t\t\t\t\t\t\t\t<div class=\"card sensors-and-communication modeling-and-simulation research-project \">\n\t<div class=\"card-img\">\n\t\t<img decoding=\"async\" width=\"250\" height=\"180\" src=\"https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/06\/Bild_Projket_Savagnini_reduziertes-Online-Modell.png\" class=\"attachment-thumbnail_389x240 size-thumbnail_389x240 wp-post-image\" alt=\"\" srcset=\"https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/06\/Bild_Projket_Savagnini_reduziertes-Online-Modell.png 250w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/06\/Bild_Projket_Savagnini_reduziertes-Online-Modell-248x180.png 248w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/06\/Bild_Projket_Savagnini_reduziertes-Online-Modell-165x119.png 165w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/06\/Bild_Projket_Savagnini_reduziertes-Online-Modell-115x83.png 115w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/06\/Bild_Projket_Savagnini_reduziertes-Online-Modell-76x55.png 76w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/06\/Bild_Projket_Savagnini_reduziertes-Online-Modell-22x16.png 22w, https:\/\/lcm.web-email.at\/\/wp-content\/uploads\/2025\/06\/Bild_Projket_Savagnini_reduziertes-Online-Modell-182x131.png 182w\" sizes=\"(max-width: 250px) 100vw, 250px\" \/>\t\t\t<\/div>\n\t<div class=\"card-body\">\n\t\t<div class=\"frame\">\n\t\t\t<h4><a href=\"https:\/\/lcm.web-email.at\/en\/automatic-sheet-metal-bending-machine\/\" rel=\"bookmark\">Automatic sheet metal bending machine<\/a><\/h4>\t\t\t<p>Digital twin in use at Salvagnini Automatic sheet metal bending machine We have developed a digital twin for predictive maintenance&#8230;.<\/p>\n\t\t<\/div>\n\t\t<a class=\"link-more\" href=\"https:\/\/lcm.web-email.at\/en\/automatic-sheet-metal-bending-machine\/\">Read more<\/a>\n\t<\/div>\n<\/div>\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t\t\t\t\t\t<div class=\"btn-center\"><a href=\"https:\/\/lcm.web-email.at\/projekte\/\"  target=\"_blank\" class=\"btn\">View more projects<\/a><\/div>\n\t\t\t\t\t<\/div>\n\t<\/section>\n","protected":false},"excerpt":{"rendered":"<p>LCM uses deep learning to precisely detect and measure hairline cracks on steel strips, even with large amounts of data and artifacts.<\/p>\n","protected":false},"author":3,"featured_media":4801,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"inline_featured_image":false,"footnotes":""},"categories":[111,142],"tags":[161,160,159,148],"class_list":["post-8398","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-sensors-and-communication","category-data-analysis-and-ai","tag-ai","tag-data-analysis","tag-image-processing","tag-smart-factory-en"],"acf":[],"_links":{"self":[{"href":"https:\/\/lcm.web-email.at\/en\/wp-json\/wp\/v2\/posts\/8398","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lcm.web-email.at\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/lcm.web-email.at\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/lcm.web-email.at\/en\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/lcm.web-email.at\/en\/wp-json\/wp\/v2\/comments?post=8398"}],"version-history":[{"count":2,"href":"https:\/\/lcm.web-email.at\/en\/wp-json\/wp\/v2\/posts\/8398\/revisions"}],"predecessor-version":[{"id":8852,"href":"https:\/\/lcm.web-email.at\/en\/wp-json\/wp\/v2\/posts\/8398\/revisions\/8852"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/lcm.web-email.at\/en\/wp-json\/wp\/v2\/media\/4801"}],"wp:attachment":[{"href":"https:\/\/lcm.web-email.at\/en\/wp-json\/wp\/v2\/media?parent=8398"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/lcm.web-email.at\/en\/wp-json\/wp\/v2\/categories?post=8398"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/lcm.web-email.at\/en\/wp-json\/wp\/v2\/tags?post=8398"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}