[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-128773-en":3,"doc-seo-128773-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},128773,1099523882367,"Hazel","https://ap-avatar.wpscdn.com/davatar_9964176cb1d06d4a9deccf72a44ae3dc",8,"Research & Report","Machine Learning for Crack Segmentation from Photogrammetric Imagery","Machine Learning for crack segmentation from photogrammetric imagery presents a research framework for detecting and quantifying cracks using deep learning and photogrammetric workflows. It introduces the motivation and objectives, reviews relevant non-destructive inspection technologies, and formulates a U-Net-based segmentation approach with dataset design, network architecture, and evaluation metrics. Case studies cover road pavement and retaining walls, including improvements to photogrammetric models via K-Means clustering, metric verification, and conversion from crack masks to point clouds.","UNIVERSITY OF SALERNO  \nDepartment of Civil Engineering  \nPhD Course  \non  \nRisk and Sustainability  \nin Civil, Architectural and Environmental Engineering Systems  \nXXXV Cycle (2019-2022)  \nMachine Learning for Crack Segmentation from Photogrammetric Imagery  \nLucas Matias Gujski  \nTutor Co-Tutor Coordinator  \nProf. Margherita Fiani  \nProf. Salvatore Barba  \nProf. Fernando Fraternali  \n[Page reserved for the Dedication]  \nSummary  \nAbstract ___________________________________________ 1  \nRiassunto _________________________________________ 2  \nResumen __________________________________________ 3  \n1. Introduction _____________________________________ 5  \n1.1 Motivation ______________________________________________ 5  \n1.2 Objectives ______________________________________________ 8  \n1.3 Outline _________________________________________________ 9  \n2. Background: Machine Learning & Deep Learning______ 11  \nInput Images ____________________________________________________ 15  \nConvolutional Layers _____________________________________________ 16  \nPooling Layers___________________________________________________ 17  \nFully Connected Layers ___________________________________________ 17  \n3. Literature review ________________________________ 21  \nVisual Inspection ________________________________________________ 22  \nUltrasonic Testing________________________________________________ 22  \nAcoustic Emission ________________________________________________ 23  \nGround-Penetrating Radar_________________________________________ 23  \nTerrestrial Laser Scanner __________________________________________ 24  \nImage-Based techniques __________________________________________ 26  \n4. U-Net-Based Crack Segmentation __________________ 35  \n4.1 Dataset________________________________________________ 36  \n4.2 Network architecture ____________________________________ 37  \n4.3 Model Evaluation Metrics_________________________________ 39  \n4.4 Results of U-Net Based Model _____________________________ 41  \n5. Methods and Applications in case studies____________ 45  \n5.1 Methods_______________________________________________ 45  \n5.1.1 Improvement of photogrammetric models: K-Means Clustering ______ 47  \nK-Means _______________________________________________________ 47  \nAccuracy Parameters _____________________________________________ 49  \nReprojection Error _____________________________________________ 49  \nAngle Between Homologous Rays_________________________________ 51  \nImage redundancy _____________________________________________ 52  \nProjection Accuracy ____________________________________________ 52  \n5.1.2 Crack metric measurements ___________________________________ 53  \nMethod for Determining Crack Area and Width Measurements ___________ 53  \n5.1.3 From the crack mask to the point cloud: Image2PointCloud _________ 54  \nImage2PointCloud _______________________________________________ 54  \n5.2 Application of the Methods in case studies ___________________ 56  \n5.2.1 Case studies ________________________________________________ 56  \nRoad pavement case study ________________________________________ 56  \nRetaining wall case study__________________________________________ 57  \nEquipment Utilized for Photogrammetric Surveys in Case Studies _________ 58  \n5.2.2 U-Net-based Model: Case Studies Results ________________________ 61  \n5.2.3 K-Means Clustering Analysis ___________________________________ 64  \n5.2.4 Metric measurements verification ______________________________ 70  \n5.2.5 Image2PointCloud application _________________________________ 76  \nConclusions and Outlook ____________________________ 81  \nAppendix _________________________________________ 85  \nReferences ______________________________________ 101  \nList of Figures  \nFigure 2.1 Diagram-Subsets ofAI ....................................................................................... 12  \nFigure 2.2 3-layer Neural Network ....................................................","cbCaifc8fC5aZKIN","https://ap.wps.com/l/cbCaifc8fC5aZKIN","pdf",18586438,1,119,"English","en",105,"# Introduction\n## Motivation\n## Objectives\n## Outline\n# Background: Machine Learning & Deep Learning\n## Input Images\n## Convolutional Layers\n## Pooling Layers\n## Fully Connected Layers\n# Literature review\n## Visual Inspection\n## Ultrasonic Testing\n## Acoustic Emission\n## Ground-Penetrating Radar\n## Terrestrial Laser Scanner\n## Image-Based techniques\n# U-Net-Based Crack Segmentation\n## Dataset\n## Network architecture\n## Model Evaluation Metrics\n## Results of U-Net Based Model\n# Methods and Applications in case studies\n## Methods\n### Improvement of photogrammetric models: K-Means Clustering\n### Crack metric measurements\n### From the crack mask to the point cloud: Image2PointCloud\n## Application of the Methods in case studies\n### Case studies\n### U-Net-based Model: Case Studies Results\n### K-Means Clustering Analysis\n### Metric measurements verification\n### Image2PointCloud application\n# Conclusions and Outlook\n# Appendix\n# References","[{\"question\":\"What is the core approach for crack segmentation in this work?\",\"answer\":\"The study builds a U-Net-based crack segmentation model using a prepared dataset, a defined network architecture, and model evaluation metrics.\"},{\"question\":\"How are photogrammetric models improved before crack analysis?\",\"answer\":\"The workflow improves photogrammetric models using K-Means clustering, supported by accuracy-related parameters such as reprojection error, projection accuracy, and related geometric measures.\"},{\"question\":\"How are crack regions used for further measurements and 3D representation?\",\"answer\":\"The method derives crack area and width measurements from the crack mask, and then converts the crack mask to a point cloud through Image2PointCloud for case-study applications.\"}]","Machine Learning for Crack Segmentation from Photogrammetric Imagery | PDF",1786003297,300,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"machine-learning-for-crack-segmentation-from-photogrammetric-imagery","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/machine-learning-for-crack-segmentation-from-photogrammetric-imagery/128773/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-06",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What is the core approach for crack segmentation in this work?","Question",{"text":75,"@type":76},"The study builds a U-Net-based crack segmentation model using a prepared dataset, a defined network architecture, and model evaluation metrics.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How are photogrammetric models improved before crack analysis?",{"text":80,"@type":76},"The workflow improves photogrammetric models using K-Means clustering, supported by accuracy-related parameters such as reprojection error, projection accuracy, and related geometric measures.",{"name":82,"@type":73,"acceptedAnswer":83},"How are crack regions used for further measurements and 3D representation?",{"text":84,"@type":76},"The method derives crack area and width measurements from the crack mask, and then converts the crack mask to a point cloud through Image2PointCloud for case-study applications.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]