[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-651112-105":3,"detail-sidebar-cat-0-en-105":80,"doc-detail-651112-en":130},{"code":4,"msg":5,"data":6},0,"ok",{"site_id":7,"language":8,"slug":9,"title":10,"keywords":11,"description":12,"schema_data":13,"social_meta":73,"head_meta":75,"extra_data":77,"updated_unix":79},105,"en","automated-dimension-aware-2d-to-bim-reconstruction-through-cross-modal-text-geometry-alignment","AUTOMATED DIMENSION-AWARE 2D-TO-BIM RECONSTRUCTION THROUGH CROSS-MODAL TEXT-GEOMETRY ALIGNMENT","","Digitising existing buildings enables energy-efficient retrofits and lifecycle analysis, yet legacy drawings often embed dimensions as text annotations, limiting metric fidelity. This paper proposes a vision-based pipeline that combines OCR, OCR-guided denoising, semantic segmentation, and grid-based alignment. Its core contribution is a cross-modal text–geometry alignment method to reconstruct a metric reference grid and connect it to building geometry. A second contribution introduces an IFC-based evaluation framework, with results showing reliable recovery of absolute dimensions for dimension-aware BIM reconstruction.",{"@graph":14,"@context":72},[15,34,55],{"@type":16,"itemListElement":17},"BreadcrumbList",[18,23,27,31],{"item":19,"name":20,"@type":21,"position":22},"https://docshare.wps.com","Home","ListItem",1,{"item":24,"name":25,"@type":21,"position":26},"https://docshare.wps.com/document/","Document",2,{"item":28,"name":29,"@type":21,"position":30},"https://docshare.wps.com/document/research-report/","Research & Report",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/automated-dimension-aware-2d-to-bim-reconstruction-through-cross-modal-text-geometry-alignment/651112/",4,{"url":32,"name":10,"@type":35,"image":36,"author":41,"headline":10,"publisher":44,"fileFormat":47,"inLanguage":8,"description":12,"dateModified":48,"datePublished":49,"encodingFormat":47,"isAccessibleForFree":50,"interactionStatistic":51},"DigitalDocument",{"url":37,"@type":38,"width":39,"height":40},"https://docshare.wps.com/thumbnails/automated-dimension-aware-2d-to-bim-reconstruction-through-cross-modal-text-geometry-alignment/651112.png","ImageObject",300,407,{"name":42,"@type":43},"cimolpedes","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-10-26","2026-10-24",true,{"@type":52,"interactionType":53,"userInteractionCount":26},"InteractionCounter",{"@type":54},"ViewAction",{"@type":56,"mainEntity":57},"FAQPage",[58,64,68],{"name":59,"@type":60,"acceptedAnswer":61},"What problem does the paper target in automated 2D-to-BIM reconstruction?","Question",{"text":62,"@type":63},"It targets the difficulty of achieving engineering-grade geometric precision when legacy 2D floorplans encode dimensions as textual annotations, where heuristic scaling or normalized coordinates often fail to preserve millimetre-level fidelity.","Answer",{"name":65,"@type":60,"acceptedAnswer":66},"What is the main methodological contribution of the proposed framework?",{"text":67,"@type":63},"The main contribution is a cross-modal text–geometry alignment strategy that reconstructs a metric reference grid from dimensional annotations and links it to component-level architectural geometry derived from segmentation.",{"name":69,"@type":60,"acceptedAnswer":70},"How does the paper evaluate reconstruction quality at engineering level?",{"text":71,"@type":63},"It introduces an IFC-based evaluation framework to assess quality at the object and parameter level, addressing the limitation of vision-oriented metrics like IoU that are insensitive to millimetre-scale deviations.","https://schema.org",{"og:url":32,"og:type":74,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":76,"canonical":32},"index,follow",{"doc_id":78,"site_id":7},651112,1792977449,{"code":4,"msg":81,"data":82},"success",[83,87,91,95,100,105,110,114,119,122,126],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":84,"show_sort_weight":85,"slug":86},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":88,"show_sort_weight":89,"slug":90},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":92,"show_sort_weight":93,"slug":94},"Exam",70,"exam",{"id":96,"doc_module":4,"doc_module_name":25,"category_name":97,"show_sort_weight":98,"slug":99},5,"Comic",60,"comic",{"id":101,"doc_module":4,"doc_module_name":25,"category_name":102,"show_sort_weight":103,"slug":104},6,"Technology",50,"technology",{"id":106,"doc_module":4,"doc_module_name":25,"category_name":107,"show_sort_weight":108,"slug":109},7,"Healthcare",40,"healthcare",{"id":111,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":112,"slug":113},8,30,"research-report",{"id":115,"doc_module":4,"doc_module_name":25,"category_name":116,"show_sort_weight":117,"slug":118},9,"Religion & Spirituality",20,"religion-spirituality",{"id":117,"doc_module":4,"doc_module_name":25,"category_name":120,"show_sort_weight":117,"slug":121},"World Cup","world-cup",{"id":123,"doc_module":4,"doc_module_name":25,"category_name":124,"show_sort_weight":123,"slug":125},10,"Lifestyle","lifestyle",{"id":127,"doc_module":4,"doc_module_name":25,"category_name":128,"show_sort_weight":96,"slug":129},19,"General","general",{"code":4,"msg":81,"data":131},{"doc_id":78,"user_id":132,"nickname":42,"user_avatar":133,"doc_module":4,"category_id":111,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":26,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":111,"language":139,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":140,"faqs":141,"seo_title":142,"seo_description":12,"update_tm":143,"read_time":117},962094243610,"https://ap-avatar.wpscdn.com/davatar_3d24733baf745e90a7e4bdd5f77d97b2","AUTOMATED DIMENSION-AWARE 2D-TO-BIM RECONSTRUCTION THROUGH CROSS-MODAL TEXT-GEOMETRY ALIGNMENT  \nXing Liang 1, Tomohiro Fukuda 1, and Nobuyoshi Yabuki2,3  \n1Division of Sustainable Energy and Environmental Engineering, Graduate School of Engineering, The  \nUniversity of Osaka, Osaka, 565-0871, Japan  \n2Advanced Research Laboratories, Tokyo City University, Setagaya-ku, Tokyo, Japan 3The University of Osaka, Suita, Osaka, Japan (Professor Emeritus)  \nAbstract  \nDigitising existing buildings supports energy-efficient retrofitting and lifecycle analysis, yet many are documented only by legacy drawings with dimensions embedded as textual annotations. Existing methods often rely on heuristic global scaling over metric fidelity. This paper presents a vision-based framework integrating optical character recognition (OCR), OCR-guided denoising, semantic segmentation, and grid-based alignment. The primary contribution is a cross-modal text–geometry alignment strategy that reconstructs a metric reference grid from annotations and links it to building geometry. A second contribution is an Industry Foundation Classes-based evaluation framework. Results demonstrate reliable recovery of absolute dimensions for dimension-aware BIM reconstruction.  \nIntroduction  \nThe building sector accounts for nearly 40% of global energy consumption (El Hafdaoui et al., 2023), highlighting the need to rapidly digitise existing building stock for energy-efficient retrofits, structural analysis, and lifecycle applications. Many advanced architecture, engineering, and construction (AEC) workflows rely on semantically rich and geometrically precise building information models (BIMs) (Schönfelder et al., 2023) . However, most existing buildings predate BIM adoption and are documented primarily by legacy two-dimensional (2D) floorplan drawings (Lu and Lee, 2017) . These drawings contain critical geometric, semantic, and dimensional information. Consequently, automated 2Dto-BIM reconstruction has become an important research topic (Lim et al., 2018) .  \nTypical 2D-to-BIM pipelines involve floorplan analysis followed by BIM instantiation. While BIM generation can often be handled deterministically once reliable inputs are available, the floorplan analysis stage remains challenging. Many existing methods prioritise topological correctness over engineering-level geometric precision, often relying on heuristic global scaling or normalised coordinates that are insufficient for millimetre-sensitive BIM reconstruction (Wang and Pajarola, 2025) .  \nAchieving engineering-grade precision requires not only detecting walls and openings, but also correctly interpreting dimensional annotations and associating them with the corresponding building geometry. This task is difficult because engineering drawings contain diverse  \ndrafting conventions, rotated or occluded texts, and annotation-induced noise (Pizarro et al., 2022) .  \nExisting studies often assume that relative geometry is sufficient or use annotations only for coarse scale estimation, without establishing explicit component-level correspondences between dimension texts and building elements (Lee et al., 2025) . Moreover, reconstruction quality is usually assessed with computer-vision-oriented metrics such as intersection over union (IoU), which are insensitive to millimetre-level geometric deviations. As a result, a model may appear visually correct while remaining inadequate for engineering use.  \nTo address these limitations, this paper proposes an automated framework for dimension-aware 2D-to-BIM reconstruction from annotated raster floorplans. Its primary methodological contribution is a cross-modal text–geometry alignment strategy that reconstructs a metric reference grid from dimensional annotations and links it to segmentation-derived architectural geometry atthe component level. Its second main contribution is an IFC-based evaluation framework for engineering-level assessment of reconstruction quality at the ","cbCaiuotSoCp0MJ8","https://ap.wps.com/l/cbCaiuotSoCp0MJ8","pdf",646813,"English","# Abstract\n# Introduction\n# Related work","[{\"question\":\"What problem does the paper target in automated 2D-to-BIM reconstruction?\",\"answer\":\"It targets the difficulty of achieving engineering-grade geometric precision when legacy 2D floorplans encode dimensions as textual annotations, where heuristic scaling or normalized coordinates often fail to preserve millimetre-level fidelity.\"},{\"question\":\"What is the main methodological contribution of the proposed framework?\",\"answer\":\"The main contribution is a cross-modal text–geometry alignment strategy that reconstructs a metric reference grid from dimensional annotations and links it to component-level architectural geometry derived from segmentation.\"},{\"question\":\"How does the paper evaluate reconstruction quality at engineering level?\",\"answer\":\"It introduces an IFC-based evaluation framework to assess quality at the object and parameter level, addressing the limitation of vision-oriented metrics like IoU that are insensitive to millimetre-scale deviations.\"}]","AUTOMATED DIMENSION-AWARE 2D-TO-BIM RECONSTRUCTION THROUGH CROSS-MODAL TEXT-GEOMETRY ALIGNMENT | PDF",1792862890]