Rough Registration of BIM Element Projection for Construction Progress Tracking

Abstract

Monitoring construction progress is essential for project management. A variety of excellent automatic schedule acquisition methods have been proposed, particularly 3D reconstruction. However, both laser-based and image-based 3D reconstruction methods rely on point clouds which have some inherent defects. In this study, a rough registration method was proposed to obtain construction progress. The method does not generate point clouds during the entire process but determines the actual construction progress by registering the target detection results and the projection of building information model elements. The method was tested at the construction sites of commercial and residential buildings. The experimental results indicated that the registration accuracy reached <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">95.13%</b> . The average external parameter calibration time and the registration time of each image are <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">18.57s</b> and <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">50.59ms</b> , respectively. Compared to similar 3D reconstruction methods, the proposed method is realistic, fast, and simple. This provides a promising method for 3D reconstruction from unordered construction-site images.