| spelling |
oapen-20.500.12657-891062024-04-03T02:23:36Z Chapter A Framework for Realistic Virtual Representation for Immersive Training Environments. Plumb, Caolan Thomas, Hannah Clark, Nigel Pour Rahimian, Farzad Pandit, Diptangshu Digital twin 3D reconstruction Virtual reality Laser scanning Photogrammetry Training simulation Unreal Engine thema EDItEUR::U Computing and Information Technology::UT Computer networking and communications::UTV Virtualization As mixed-reality (XR) technology becomes more available, virtually simulated training scenarios have shown great potential in enhancing training effectiveness. Realistic virtual representation plays a crucial role in creating immersive experiences that closely mimic real-world scenarios. With reference to previous methodological developments in the creation of information-rich digital reconstructions, this paper proposes a framework encompassing key components of the 3D scanning pipeline. While 3D scanning techniques have advanced significantly, several challenges persist in the field. These challenges include data acquisition, noise reduction, mesh and texture optimisation, and separation of components for independent interaction. These complexities necessitate the search for an optimised framework that addresses these challenges and provides practical solutions for creating realistic virtual representations in immersive training environments. The following exploration acknowledges and addresses challenges presented by the photogrammetry and laser-scanning pipeline, seeking to prepare scanned assets for real-time virtual simulation in a games-engine. This methodology employs both a camera and handheld laser-scanner for accurate data acquisition. Reality Capture is used to combine the geometric data and surface detail of the equipment. To clean the scanned asset, Blender is used for mesh retopology and reprojection of scanned textures, and attention given to correct lighting details and normal mapping, thus preparing the equipment to be interacted with by Virtual Reality (VR) users within Unreal Engine. By combining these elements, the proposed framework enables realistic representation of industrial equipment for the creation of training scenarios that closely resemble real-world contexts 2024-04-02T15:46:44Z 2024-04-02T15:46:44Z 2023 chapter ONIX_20240402_9791221502893_75 2704-5846 9791221502893 https://library.oapen.org/handle/20.500.12657/89106 eng Proceedings e report application/pdf n/a 9791221502893_26.pdf https://books.fupress.com/doi/capitoli/979-12-215-0289-3_26 Firenze University Press 10.36253/979-12-215-0289-3.26 10.36253/979-12-215-0289-3.26 bf65d21a-78e5-4ba2-983a-dbfa90962870 9791221502893 137 14 Florence open access
|
| description |
As mixed-reality (XR) technology becomes more available, virtually simulated training scenarios have shown great potential in enhancing training effectiveness. Realistic virtual representation plays a crucial role in creating immersive experiences that closely mimic real-world scenarios. With reference to previous methodological developments in the creation of information-rich digital reconstructions, this paper proposes a framework encompassing key components of the 3D scanning pipeline. While 3D scanning techniques have advanced significantly, several challenges persist in the field. These challenges include data acquisition, noise reduction, mesh and texture optimisation, and separation of components for independent interaction. These complexities necessitate the search for an optimised framework that addresses these challenges and provides practical solutions for creating realistic virtual representations in immersive training environments. The following exploration acknowledges and addresses challenges presented by the photogrammetry and laser-scanning pipeline, seeking to prepare scanned assets for real-time virtual simulation in a games-engine. This methodology employs both a camera and handheld laser-scanner for accurate data acquisition. Reality Capture is used to combine the geometric data and surface detail of the equipment. To clean the scanned asset, Blender is used for mesh retopology and reprojection of scanned textures, and attention given to correct lighting details and normal mapping, thus preparing the equipment to be interacted with by Virtual Reality (VR) users within Unreal Engine. By combining these elements, the proposed framework enables realistic representation of industrial equipment for the creation of training scenarios that closely resemble real-world contexts
|
