Research Projects
As Primary Investigator (PI)/Co-Investigator (Co-I)
Scan-to-BIM Automation System for Built Assets Digitalization in Hong Kong
Innovation and Technology Fund (ITF) Tier-1 (ITP/004/23LP), HKD 7,510,000 Duration: 4/2023 - 3/2025 Status: Ongoing Role: Co-IKeywords: 3D point cloud; Deep learning; BIM; CIM; Open BIM.
Assessing Human-perceived Window View Openness in High-rise High-density Cities: An Automatic Machine Learning-based City Information Modeling Approach
HKU Leung Kau Kui and Run Run Shaw Research and Teaching Endowment Funds, HKD 50,000 Duration: 1/2023 - 12/2023 Status: Ongoing Role: PIAbstract: High window view openness benefits urban dwellers’ health and well-being, especially in high-rise, high-density cities. The benefits such as stress relief and mood restoration are further amplified for inhabitants in narrow and small rooms, especially in the post-Covid-19 era. However, the crowded cityscapes and vertical development of the urban environment lead to an imbalanced sharing of window view openness. Currently, there still exists no consensus on the definition of human-perceived window view openness in high-rise, high-density cities, and the assessment methods are limited to small-scale sites and inaccurate. Thus, an urban-scale accurate window view openness assessment is significant in examining the disparity of openness possession and providing quantified evidence for precise decision-making in the healthy high-rise, high-density urban development. The objectives of the proposed project include: i) To develop a human-perceived window view openness index using a hierarchy of window view openness characteristics. ii) To propose an urban-scale assessment method for human-perceived window view openness using photorealistic CIM and Automatic Machine Learning (AutoML). iii)To visualize urban-scale human-perceived window view openness using a 3D GIS platform for the evaluation, planning, and design of healthy high-rise, high-density urban development.
From 3D Real Scene to 3D Semantics: Reconstruction of Semantic Volumetric Building Models using 3D Skeletons in Urban Point Clouds
Natural Science Fund of Guangdong (2023A1515010757), HKD 121,000 Duration: 1/2023 - 12/2025 Status: Ongoing Role: Co-IKeywords: Building skeleton; Volumetric reconstruction; Semantic analytics