Towards Automated Retrieval of 3D Designed Data in 3D Sensed Data

摘要

Construction productivity, one major project performance indicator, requires the continuous assessment of construction progress over time. Current practices for construction progress assessment are labour intensive, expensive, and generally result in partial and sometimes erroneous information. It is thus difficult to make appropriate and timely management decisions. However, recent three-dimensional (3D) Computer Aided Design (CAD), 3D scanning and global positioning technologies present an opportunity for developing more efficient, robust and automated approaches. After reviewing and analyzing previous works conducted in the field of automated object detection, this paper presents a new approach taking advantage of global positioning technologies for robustly retrieving 3D CAD elements within 3D scanned data. This approach is implemented and experimented in laboratory. Results are conclusive towards further experiments for its application to many different fields including automated construction progress assessment for effortless productivity tracking.