Towards Object Driven Floor Plan Extraction From Laser Point Cloud

Abstract

ABSTRACT:During the last years, the demand for indoor models has increased for various purposes. As a provisional step to proceed towards higher dimensional indoor models, powerful and flexible floor plans can be utilised. Therefore, several methods have been proposed that provide automatically generated floor plans from laser point clouds. The prevailing methodology seeks to attain semantic enhancement of a model (e.g. the identification and labelling of its components) built upon already reconstructed (a priori) geometry. In contrast, this paper demonstrates preliminary research on the possibility to directly incorporate semantic knowledge, which is itself derived from the raw data during the extraction, into the geometric modelling process. In this regard, we propose a new method to automatically extract floor plans from raw point clouds. It is based on a hierarchical space partitioning of the data, integrated with primitive selection actuated by object detection. First, planar primitives corresponding to vertical architectural structures are extracted using M-estimator SAmple and Consensus (MSAC). The set of the resulting line segments are refined by a selection process through a novel door detection algorithm, considering optimization of prior information and fitness to the data. The selected lines are used as hyperlines to partition the space into enclosed areas. Finally, a floor plan is extracted from these partitions by Minimum Description Length (MDL) hypothesis ranking. The algorithm is applied on a real mobile laser scanner dataset and the results are evaluated both in terms of door detection and consecutive floor plan extraction.

Hierarchical Topometric Representation of 3D Robotic Maps

“…However, this method requires the scan data to include pose as input, which is unavailable in most of the open-source datasets. Babacan et al (2016) get the 2D floor plan by slicing the 3D point cloud at a height slightly below the ceiling, to avoid the influence of most furniture items. Then it uses door detection by projecting points to the 2D plane and wall plane detection with RANSAC.…”

“…Other, often similar segmentation algorithms have been presented Wurm et al (2008); Birk (2015, 2013) to improve the accuracy of map segmentation. Many segmentation and geometric structure extraction methods are proposed Babacan et al (2016); Ochmann et al (2019) for 3D point clouds. However, only a few of them concentrate on topology extraction.…”

Building Floor Plan Reconstruction From Slam-Based Point Cloud Using Ransac Algorithm

“…Anagnostopoulos et al (2016) offered a stable method of extraction walls, ceilings and floors that automatically processes point cloud and identifies and categorizes complication. Simultaneously, Babacan et al (2016) proposed a new method for automatically extracting floor plans by door detection algorithm from raw laser scanner. Recently, Giorgini et al (2018) presented a new approach to automate the production of the floor plan from point cloud captured by a laser scanners.…”

“…Therefore, researchers have been focusing on developing automated techniques in order to obtain interior geometric details of buildings. The aim of developing new processing methods are automatic feature extraction and model generation from point cloud, which should be able to process high-density data (Babacan et al, 2016).…”

Furniture Free Mapping using 3D Lidars

“…The approach in [16] utilized random sample consensus (RANSAC) and plane normal orientations to extract structures of the building, such as floor, ceiling, and walls. The study of Babacan [17] used minimum description length (MDL) hypothesis ranking to extract the floor blueprint. Oesau [18] presented an energy minimization function for different segment rooms in 2d maps.…”