Multi-Building Extraction and Alignment for As-Built Point Clouds: A Case Study With Crane Cameras

Abstract

Automatic reality capture and monitoring of construction sites can reduce costs, accelerate timelines and improve quality in construction projects. Recently, automatic close-range capture of the state of large construction sites has become possible through crane and drone-mounted cameras, which results in sizeable, noisy, multi-building as-built point clouds. To infer construction progress from these point clouds, they must be aligned with the as-designed BIM model. Unlike the problem of aligning single buildings, the multi-building scenario is not well-studied. In this work, we address some unique issues that arise in the alignment of multi-building point clouds. Firstly, we show that a BIM-based 3D filter is a versatile tool that can be used at multiple stages of the alignment process. We use the building-pass filter to remove non-building noise and thus extract the buildings, delineate the boundaries of the building after the base is identified and as a post-processing step after the alignment is achieved. Secondly, in light of the sparseness of some buildings due to partial capture, we propose to use the best-captured building as a pivot to align the entire point cloud. We propose a fully automated three-step alignment process that leverages the simple geometry of the pivot building and aligns partial xy-projections, identifies the base using z-histograms and aligns the bounding boxes of partial yz-projections. Experimental results with crane camera point clouds of a large construction site show that our proposed techniques are fast and accurate, allowing us to estimate the current floor under construction from the aligned clouds and enabling potential slab state analysis. This work contributes a fully automated method of reality capture and monitoring of multi-building construction sites.

Automated Computer Vision-Based Construction Progress Monitoring: A Systematic Review

“…Various studies have employed these systems for element recognition, 3D point cloud generation, and progress calculation [41,72,73]. Few studies have also mounted these systems on a crane to cover a large area and provided less occluded 3D point clouds for construction progress estimation [33,74]. One study [74] addressed the technical challenge of multi-building extraction and alignment of as-built point clouds.…”

“…Few studies have also mounted these systems on a crane to cover a large area and provided less occluded 3D point clouds for construction progress estimation [33,74]. One study [74] addressed the technical challenge of multi-building extraction and alignment of as-built point clouds. The study utilized the data captured through two stereo cameras installed on a tower crane on a mi-used construction project including a shopping mall, hotel, housing, and offices.…”

“…There are two distinct methods of image model registration: coarse registration and fine registration. The coarse registration along with post-processing allows for rough alignment, whereas the fine registration can achieve near-optimal alignment [74]. The coarse registration can be achieved through various approaches, such as plane-based matching, principal component analysis-based alignment [47], plane patches-based matching, 3D to 2D transformation [119], building extraction, and alignment for multi-building point clouds [74], etc.…”

“…The coarse registration along with post-processing allows for rough alignment, whereas the fine registration can achieve near-optimal alignment [74]. The coarse registration can be achieved through various approaches, such as plane-based matching, principal component analysis-based alignment [47], plane patches-based matching, 3D to 2D transformation [119], building extraction, and alignment for multi-building point clouds [74], etc. For example, [84] proposed a semi-automated plane-based coarse registration approach and compared the proposed method with already existing general-purpose registration software, and reduced the complexities and time requirements associated with this process.…”

“…However, the plane patchedbased system is the state-of-art of current practice, allowing for automatic registration using a 4-plane approach rather than a 3-plane as in a plane-based process [85]. Moreover, the Iterative Closest Point (ICP) is most frequently used for fine registration in CV-based CPM studies [47,74,84,119]. For example, [120] proposed a fully automated registration of 3D data to a 3D CAD model for CPM purposes.…”