"BIM-to-Scan" for Scan-to-BIM: Generating Realistic Synthetic Ground Truth Point Clouds based on Industrial 3D Models

Abstract

In the field of Scan-to-BIM, recent developments achieve promising results in accuracy and flexibility, leveraging tools from the field of deep learning for semantic segmentation of raw point cloud data. Those methods demand large-scale, domain-specific datasets for training. Promising ideas to fulfill this need use primitive synthetic point cloud data, which predominantly lack distinct point cloud properties, such as missing patches due to occlusions in the scene. To solve this issue, we use a specialized laser scan simulation tool from the domain of Geosciences in a toolchain that allows generating realistic ground truth data based on 3D models. In this context, we introduce a comprehensive taxonomy for the industrial point cloud context. Furthermore, we provide the missing link for a comprehensive, open-source toolchain that is flexible towards any use case in the field.

Virtual LiDAR Simulation as a High Performance Computing Challenge: Toward HPC HELIOS++

“…Furthermore, expected scenes from LiDAR sensing can be of very different types. For instance, works using HELIOS++ (or its predecessor HELIOS for Java) range from industrial applications [8] to forestry applications [9]. Scenarios of real laser scanning range from mobile laser scanning (MLS) for autonomous driving [10] or indoor mapping [11], to This article has been accepted for publication in IEEE Access.…”

“…For this purpose, let L(i) = |{x ∈ H * : x < b i }| be the number of primitives on left partition if split is placed on i-th bin and R(i) = |{x ∈ H * : x ≥ b i }| its analogous for right partition. Now, finding the approximated optimum split position r opt can be done by solving (8). This computation has O(m) complexity.…”

“…Despite this, a priori finding a sub-optimal but good enough warehouse factor is not too complicated for the general case. Empirical evidence from Section V-A and Section V-C shows that f ∈ [4,8] and m ∈ [32,64] are good choices for different scenes.…”

Parametrisierte Modellierung für den Einsatz von KI am Beispiel Betonbrückenbau

“…erfasst werden. Da die Erfassung sehr zeitintensiv ist und wenige Daten zur Verfügung stehen, können synthetische Daten Abhilfe leisten [7]. Griffiths…”

“…erfasst werden. Da die Erfassung sehr zeitintensiv ist und wenige Daten zur Verfügung stehen, können synthetische Daten Abhilfe leisten [7]. Griffiths et al [8] haben für die Bereitstellung von Punktwolken virtuelle 3D‐Räume genutzt, die zu Zwecken der Visualisierung frei im Internet zugänglich sind.…”