Abstract
This paper describes a mobile application that builds and updates a 3D model of an indoor environment, including walls, floor and openings, by a simple scan performed using a tablet equipped with a depth sensor. This algorithm is fully implemented on the device, does not require internet connection and runs in real-time, i.e., at five frames per second. This is made possible by taking advantage of recent AR frameworks, by assuming that the structure of the room is aligned on an Euclidean grid and by simply starting the scan in front of a wall. The wall detection is achieved in two steps. First, each incoming point cloud is segmented into planar wall candidates. Then, these planes are matched to the previously detected planes and labeled as ground, ceiling, wall, openings or noise depending on their geometric characteristics. Our evaluations show that the algorithm is able to measure a plane-to-plane distance with a mean error under 2 cm, leading to an accurate estimation of a room dimensions. By avoiding the generation of an intermediate 3D model, as a mesh, our algorithm allows a significant performance gain. The 3D model can be exported to a CAD software, in order to plan renovation works or to estimate energetic performances of the rooms. In the user experiments, a good usability score of 75 is obtained.
Highlights
Creating a 3D model of an existing building has found many applications, such as the generation of a BIM1 of the building or obtaining geometrical information about the building
The maximum error is obtained for the meeting room, where the ceiling lights distort the measurements of the depth sensor
This paper has described an application that runs on a tablet equipped with a depth sensor that generates and updates a 3D model of an indoor environment
Summary
Creating a 3D model of an existing building has found many applications, such as the generation of a BIM1 of the building or obtaining geometrical information about the building (dimensions, surfaces, etc.,). With the recent release of depth sensors integrated onto tablets and with the enhancement of their computing capabilities, it is possible to use these devices to perform a real-time 3D reconstruction. The proposed algorithm uses the computing capabilities of the device to generate a 3D editable model on-the-fly, including the walls and openings of the rooms. A planar segmentation is performed in each depth image of the sequence and the extracted planes are matched to the previously detected ones Through this temporal analysis, walls are extracted and described in terms of geometry and, the global 3D model is updated. This paper explains the different components of a more comprehensive mobile application, which is able to create and export a 3D model that can be used and edited in a CAD software, to plan renovation works and to estimate energetic performance.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
