Towards Automated Measurement of As-Built Components Using Computer Vision

Area of Science:

• Construction Engineering and Management
• Computer Vision
• Artificial Intelligence

Background:

• Accurate construction progress monitoring is essential for project adherence to timelines and budgets.
• Current methods like 3D laser scanning and photogrammetry are complex and challenging to deploy on dynamic sites.
• Frequent physical site observations are needed to verify construction progress.

Purpose of the Study:

• To develop an automated pipeline for measuring as-built components using AI and computer vision.
• To overcome the limitations of existing 3D measurement techniques in construction environments.
• To enable accurate size measurement of construction elements from a single image.

Main Methods:

• Proposed a novel pipeline utilizing artificial intelligence and computer vision.
• Employed a stereo camera system to capture a single image for data acquisition.
• Developed a method for automated measurement of as-built components like concrete walls and columns.

Main Results:

• Successfully demonstrated the automated measurement of concrete wall and column sizes.
• Achieved accurate geometric property ascertainment using a single image and target.
• Validated a fully automated computer vision-based measurement method.

Conclusions:

• The proposed pipeline offers a simplified and automated approach to measuring as-built components.
• This method is suitable for measuring components in built assets and has potential for progress monitoring.
• Integration with Building Information Modelling (BIM) applications can further enhance construction project management.