Skip to content Skip to navigation

Evaluation of Computer Vision-and 4D BIM-based Construction Progress Tracking on a UAV Platform

TitleEvaluation of Computer Vision-and 4D BIM-based Construction Progress Tracking on a UAV Platform
Publication TypeTechnical Report
Year of Publication2017
AuthorsHamledari, H, McCabe, B, Davari, S, Shahi, A, Azar, E, Flager, F
IssueTR235
Publisher6TH CSCE/ASCE/CRC INTERNATIONAL CONSTRUCTION SPECIALTY CONFERENCE
Keywords4D BIM, Automation, Computer Vision, construction progress tracking, digital images, Unmanned Aerial Vehicles
Abstract

The application of image-based progress tracking and object detection techniques has recently been extended to dynamic automated data collection and image capture platforms such as unmanned aerial vehicles (UAV). The use of UAVs has a great potential to eliminate tedious, labor-intensive, costly, and manual image capture processes. It can also provide a clearer and more informative view of construction work due to the UAVs’ high agility and maneuverability. However, it is also of utmost importance to analyze the effect of UAVs’ highly dynamic behavior on the accuracy of image-based solutions. UAV-captured images are subject to motion blur which not only can jeopardizes the object and progress recognition accuracy, but also the quality and reliability of resulting as-built 4D building information models (BIM). This study evaluates the performance of a 4D BIM-and computer vision-based construction progress detection method on images captured by an unmanned aerial vehicle. In this research, the components of indoor partitions such as studs, insulation, electrical outlets, and state of drywall work are automatically detected, and the 4D BIM is updated with schedule and progress information. In a series of experiments, the accuracy of this solution is analyzed with respect to the UAV’s velocity and photo capture configuration. This analysis can benefit UAV-based progress tracking systems and facilitate reliable UAV-based data collection at construction sites.

URLhttps://purl.stanford.edu/wh873cw2351
Citation Key1576