| Author: | Jin, Yixuan |
| Title: | Spatial augmented reality for 3-Dimensional (3D) architectural design visualization and collaboration |
| Advisors: | Seo, JoonOh (BRE) |
| Degree: | Ph.D. |
| Year: | 2025 |
| Department: | Department of Building and Real Estate |
| Pages: | xii, 126 pages : color illustrations |
| Language: | English |
| Abstract: | As building Information modelling (BIM) and augmented reality (AR) becomes more commonplace in collaborative design, a number of critical shortcomings have emerged--most noticeably in the areas of technical performance, user engagement, and overall visual quality. These challenges have, in turn, stimulated renewed curiosity in Spatial Augmented Reality (SAR) approaches. Unlike conventional mixed reality approaches, SAR technology projects digital content directly onto physical surfaces, achieving an unprecedented seamless integration of virtual and real environments with photorealistic quality. This advanced approach provides superior immersion, spatial cognition, and visualization fidelity compared to both immersive virtual reality systems and traditional augmented reality implementations. Conventional AR systems primarily operate through direct 3D-to-3D alignment, where digital models are matched precisely with their physical counterparts. SAR, however, employs a more complex workflow - first converting 3D data into 2D projections that are then remapped into 3D space. This dimensional transformation presents significant technical hurdles in preserving both geometric precision and visual quality throughout the conversion process. The challenges become particularly pronounced when simulating daylight effects, where SAR systems must use stationary projectors to represent the sun's dynamic movement, which is a stark contrast to BIM environments that can directly simulate solar positions. These technical considerations are especially critical for architectural visualization, where accurately representing building exteriors and environmental impacts remains paramount for effective design collaboration. The research framework consequently focuses on addressing these fundamental challenges across both domains. This research firstly presents an BIM-based projection mapping framework that utilizes SAR technology to enhance architectural design collaboration. The comprehensive framework details the complete procedural workflow and technical specifications for generating 3D models and applying textures to physical surfaces. Through rigorous testing with a projection mapping tool, we successfully demonstrated stable projection of virtual building models onto physical objects, effectively addressing key limitations of existing AR-based collaboration systems. Our experimental results reveal significant improvements from optimized control parameters, including enhanced model clarity at larger scales and superior stereoscopic projection in low-light conditions. However, persistent challenges with manual calibration for complex geometries and dynamic projection scenarios highlight the need for automated calibration solutions to fully unlock SAR's potential for cross-disciplinary collaboration. To overcome these calibration challenges, we developed an automated geometric calibration system employing a projector-camera configuration. Our innovative solution utilizes machine vision algorithms operating with an uncalibrated projector-camera pair to establish precise spatial alignment. Extensive validation across diverse building models and hardware setups confirmed the system's high accuracy. The solution offers multiple advantages, including the elimination of separate projector-camera calibration requirements, a streamlined single-phase calibration process, simplified user interaction requiring only white panel placement and image capture, and seamless integration with BIM model textures for enhanced design compatibility. The research further extends to daylight simulation applications in architectural design, addressing the realism and collaboration limitations inherent in current mixed reality technologies. Our SAR-based daylight simulation system incorporates advanced solar position algorithms and a custom-developed plugin for the SPARK platform, generating high-fidelity building textures, daylight beams, and shadow patterns with exceptional accuracy. Comparative studies demonstrate SAR's superior performance over software-based simulations across critical metrics, including visual authenticity, spatial presence, and user engagement. By enabling real-time feedback and immersive visualization capabilities, the system facilitates a more dynamic and interactive design process accessible to both professional practitioners and public stakeholders. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/13921