Kinetic Shape Reconstruction

Abstract

Converting point clouds into concise polygonal meshes in an automated manner is an enduring problem in Computer Graphics. Prior work, which typically operate by assembling planar shapes detected from input points, largely overlooked the scalability issue of processing a large number of shapes. As a result, they tend to produce overly simplified meshes with assembling approaches that can hardly digest more than one hundred shapes in practice. We propose a shape assembling mechanism which is at least one order magnitude more efficient, both in time and in number of processed shapes. Our key idea relies upon the design of a kinetic data structure for partitioning the space into convex polyhedra. Instead of slicing all the planar shapes exhaustively as prior methods, we create a partition where shapes grow at constant speed until colliding and forming polyhedra. This simple idea produces a lighter yet meaningful partition with a lower algorithmic complexity than an exhaustive partition. A watertight polygonal mesh is then extracted from the partition with a min-cut formulation. We demonstrate the robustness and efficacy of our algorithm on a variety of objects and scenes in terms of complexity, size and acquisition characteristics. In particular, we show the method can both faithfully represent piecewise planar structures and approximating freeform objects while offering high resilience to occlusions and missing data.

Floorplan generation from 3D point clouds: A space partitioning approach

“…To tackle this problem, we project wall planes onto the X-Y plane and partition the 2D space using a kinetic datastructure described in [35] (as illustrated in Figure 2d). As explained in [36],…”

Efficient Lightweight Surface Reconstruction Method from Rock-Mass Point Clouds

“…In addition, the constructed model might be made erroneous due to the high corruption of inputted data. On this basis, Bauchet et al [33] proposed a dynamic reconstruction method. Rather than decomposing the space completely, it adopts gradual extension.…”

Optimal Model Fitting for Building Reconstruction From Point Clouds

“…Reference [18] presented an improved RANSAC algorithm to avoid abnormal and/or infinite solutions which are typically encountered in previously published methods that use the rooftop primitive adjacency matrix to solve the critical rooftop vertices. To assemble a number of detected planar shapes from point clouds, Bauchet and Lafarge [19] presented an efficient shape assembling mechanism with kinetic data structure to covert point cloud into watertight polygon meshes, and demonstrated its efficiency on large-scale downtown buildings reconstruction, while it relied on standard planar shape detection algorithms and was limited to roof reconstruction with planar shapes. To overcome the problems of outliers and/or noise in RANSAC, reference [20] presented a PCA-based robust segmentation algorithm for laser scanning 3D point cloud.…”

“…In terms of model representation, the outputs of most existing methods are either polygonal meshes [19], [31] or parameterized building models [32], both are geometric models without semantic information, which limit the potential application of these models.…”