BALTIMORE - March 16, 2023 - PRLog -- By automatically identifying damage in photos, computer vision has the potential to help with building inspection after an earthquake. The worldwide impact of the damage on the structural system must be considered while determining safety of a building damaged by an earthquake. While conducting an inspection, it is important to consider the connected component's projected damage development and performance-related role in the structural system.
The Building Information Modeling is used as a reference platform in the digital twin framework for post-earthquake building evaluation that includes unmanned aerial vehicle (UAV) imaging, component identification, and damage evaluation. If damage is found, a GrabCut-based segmentation technique is used to allocate each image pixel to a particular BIM component. In order to identify nonstructural damage and link it to certain BIM components, 3D point cloud change detection is used.
Instances:
BIM-guided image selection and component identification while creating a digital twin (https://www.tejjy.com/digital-twin-building/) for an existing reinforced concrete frame structure.
Identifying damaged nonstructural brick walls using 3D point cloud change detection in a synthetic graphics environment.
Sukh Singh, V.P.,Tejjy BIM Inc mentioned – "To anticipate building's post-earthquake safety based on UAV survey, the framework combines component-wise damage estimates with pre-earthquake structural analysis."
Development Process of Digital Twin Framework:
Getting an overview of digital twin's parts - Before earthquake, a preliminary evaluation of the structure, comprising a walkthrough, sketch review, and structural analysis, serves as the first step in creating the digital twin.
Implementing software: Next is using software and BIM of the as-built building shell is created.
Creating 3D point cloud of structure: A preliminary photographic survey of the structure is then flown, and a 3D point cloud of the target structure is made and synchronized with the BIM.
Building digital twin: The digital twin is made up of the aligned BIM and point cloud, collection of survey photos, and structural model analysis.
Producing 3D point cloud: An updated 3D point cloud is produced from the survey photos following the completion of a second, post-earthquake UAV scan.
Spotting building elements: The digital twin directs the survey's image selection process and helps to spot building elements in the photographs.
Using 3D change detection: The digital twin also uses 3D change detection to distinguish damaged nonstructural components from point clouds taken before and after the earthquake.
Connecting damage to building components: By connecting damage to specific building components, each component's projected performance is considered to deduce component-level requirements and subsequently enhance building level performance and safety.
Account Phone Number