TY - JOUR A1 - Artus, Mathias A1 - Alabassy, Mohamed Said Helmy A1 - Koch, Christian T1 - A BIM Based Framework for Damage Segmentation, Modeling, and Visualization Using IFC JF - Applied Sciences N2 - Paper-based data acquisition and manual transfer between incompatible software or data formats during inspections of bridges, as done currently, are time-consuming, error-prone, cumbersome, and lead to information loss. A fully digitized workflow using open data formats would reduce data loss, efforts, and the costs of future inspections. On the one hand, existing studies proposed methods to automatize data acquisition and visualization for inspections. These studies lack an open standard to make the gathered data available for other processes. On the other hand, several studies discuss data structures for exchanging damage information among different stakeholders. However, those studies do not cover the process of automatic data acquisition and transfer. This study focuses on a framework that incorporates automatic damage data acquisition, transfer, and a damage information model for data exchange. This enables inspectors to use damage data for subsequent analyses and simulations. The proposed framework shows the potentials for a comprehensive damage information model and related (semi-)automatic data acquisition and processing. KW - Building Information Modeling KW - Brücke KW - Inspektion KW - Maschinelles Lernen KW - Bildverarbeitung KW - Building Information Modeling KW - Bridge KW - Inspection KW - Damage Segmentation KW - Machine Learning KW - OA-Publikationsfonds2022 Y1 - 2022 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:gbv:wim2-20220314-46059 UR - https://www.mdpi.com/2076-3417/12/6/2772 VL - 2022 IS - volume 12, issue 6, article 2772 SP - 1 EP - 24 PB - MDPI CY - Basel ER - TY - THES A1 - Alabassy, Mohamed Said Helmy T1 - Automated Approach for Building Information Modelling of Crack Damages via Image Segmentation and Image-based 3D Reconstruction N2 - As machine vision-based inspection methods in the field of Structural Health Monitoring (SHM) continue to advance, the need for integrating resulting inspection and maintenance data into a centralised building information model for structures notably grows. Consequently, the modelling of found damages based on those images in a streamlined automated manner becomes increasingly important, not just for saving time and money spent on updating the model to include the latest information gathered through each inspection, but also to easily visualise them, provide all stakeholders involved with a comprehensive digital representation containing all the necessary information to fully understand the structure’s current condition, keep track of any progressing deterioration, estimate the reduced load bearing capacity of the damaged element in the model or simulate the propagation of cracks to make well-informed decisions interactively and facilitate maintenance actions that optimally extend the service life of the structure. Though significant progress has been recently made in information modelling of damages, the current devised methods for the geometrical modelling approach are cumbersome and time consuming to implement in a full-scale model. For crack damages, an approach for a feasible automated image-based modelling is proposed utilising neural networks, classical computer vision and computational geometry techniques with the aim of creating valid shapes to be introduced into the information model, including related semantic properties and attributes from inspection data (e.g., width, depth, length, date, etc.). The creation of such models opens the door for further possible uses ranging from more accurate structural analysis possibilities to simulation of damage propagation in model elements, estimating deterioration rates and allows for better documentation, data sharing, and realistic visualisation of damages in a 3D model. KW - Building Information Modeling KW - BIM KW - IFC KW - Damage Information Modelling KW - Cracks Segmentation KW - Cracks 3D Modelling KW - Netscape Internet Foundation Classes Y1 - 2021 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:gbv:wim2-20230818-64162 ER -