No abstract

No abstract

The three-dimensional mapping of the built environment is of particular importance for engineering applications such as monitoring work-in-progress

The mapping of real-world objects to 3D geometry is of particular importance for engineering applications such as as-built modeling for progress monitoring and energy performance simulation. The state-of-the-art methods for fitting NURBS to point clouds still fail to account for all the topological variations or struggle with the mapping of physical space to parameter space given unordered, incomplete, and noisy point clouds. To address these limitations, we present a new method which starts by successively fitting uniform B-spline curves in 2D as planar cross sectional cuts on the surface. An intermediate B-Spline surface is then computed by globally optimizing the cross sections and lofting over the cross sections. This surface is used to parameterize the points and perform final refinement to a NURBS surface. Assuming an input of points that can be described by a single, non-self-intersecting NURBS surface, the algorithm outputs a NURBS surface. Experimental results on several real-world point clouds show the applicability of the proposed method for asbuilt modeling of civil infrastructure systems.

Accurate and rapidly produced 3D models of the as-built environment can be significant assets for a variety of Civil Engineering scenarios. Starting with a point cloud of a scenegenerated using laser scanners or image-based reconstruction method-the user must first identify collections of points that belong to individual surfaces, and then, fit surfaces and solid geometry objects appropriate for the analysis. When performed manually, this task is often prohibitively time consuming and, in response, several research groups have recently focused on developing methods for automating the modeling process. Due to the limitations of the data collection processes as well as the complexity of as-built scenes, automated 3D modeling still presents many challenges. To overcome existing limitations, in this paper, we propose a new region growing method for robust context-free segmentation of unordered point clouds based on geometrical continuities. In our method, only one parameter is required to be set by the user to account for the desired level of abstraction. Preliminary experimental results from two challenging scenes of the built environment demonstrate that our method can account for variability in point cloud density, surface roughness, curvature, and clutter within a single scene. INTRODUCTION3D modeling of the as-built environment is used by the AEC industry in a variety of engineering analysis scenarios. Significant applications include progress monitoring of construction sites, quality control of fabrication and on-site assembly, energy performance assessment, and structural integrity evaluation. In recent years, point clouds have become the predominant data type collected on site and used as a basis for modeling. This data can originate from image-based 3D reconstruction methods using images or videos, as well as structured light methods, mainly laser scanners. The process of generating 3D models from point cloud data involves two steps: 1) identifying collections of points that belong to each surface, and 2) fitting geometry (meshes, primitives, NURBS, subdivision) to them. This process is manual and very time-consuming. In response, several research groups (e.g., Xiong et al. 2013, Zhang et al. 2013 have recently focused on developing methods for automated modeling.Although, the type of geometry to be fitted is heavily dependent on a specific analysis, the common task in Civil Engineering scenarios is the segmentation of point clouds into identifiable surfaces. Figure 1 shows examples of the use of automated segmentation to identify specific elements that need to be modeled. The complexity of as-built scenes produces significant challenges for automated segmentation: 1) Density: Point cloud models exhibit locally variable densities based on orientation and distance from the capture device. Furthermore, occlusions from surface irregularities and adjacent objects produce regions with missing data.