Computer Vision Approach for Indoor Location Recognition Within an Augmented Reality Mobile Application
PubDate: October 2019
Teams: Free University of Bolzano
Writers: Gabriele MinneciAlice SchweigkoflerCarmen MarcherGabriele Pasetti MonizzaTammam TilloDominik T. Matt
Abstract
The Building Industry (BI) lacks in sharing information among different and heterogeneous stakeholders. With the advent of new technologies such as mobile devices, IoT and Industry 4.0 different approaches have been introduced, but it is still an open research field. The present essay aims at moving a step forward using augmented reality connected with the Building Information Modelling (BIM) methodology enhancing the effectiveness of process in the BI.
The authors improved the Fraunhofer Italia’s AR4Construction (AR4C) mobile application, that lets construction workers visualize the BIM model of a building on site in augmented reality and interact with its design information. The main purpose of this study consists in studying and developing an indoor location functionality for the AR4C application. The research focuses on understanding the starting point and the orientation of the user, through the application of 3D mathematical structures and computer vision methods.
The result is a core module for an indoor location functionality. We have evaluated distances and orientation errors, analyzing the results according to the implementation strategy. In conclusion, possible improvements have been analyzed within this paper, which offers a first approach for visualization-only based frameworks aiming at indoor location and navigation and using BIM models as a starting point.
