Image-based Indoor Localisation and Navigation with wearable AR devices

Gaurav Gupta, Nishant Kejriwal, Prasun Pallav, Ehtesham Hassan, Swagat Kumar, Ramya Hebbalaguppe

Research output: Contribution to journalArticle

Abstract

We present a novel image-based indoor localisation and navigation application framework with augmented reality devices, specifically android based wearable devices. Our framework provides a robust location and navigation features in an unknown indoor environment using AR device as the visual sensor. We have used Google Glass(GG) for for conducting experiments and to demonstrate proof of concept. The GG’s limited computational power is overcome by performing all computations at back-end; GG is only used for communicating with back-end and as a sensing and display device. We use visual odometry and inertial sensors for dynamic map creation. We use SURF[1] with Bag of Visual Words(BoVW)[14] aggregation for image matching to find loop closure for revisited landmarks in the scene. The application framework generates a topological map at run-time and displays it on GG screen. The potential use cases of our work include the virtual tours in an industry setting, museums and art galleries, and research facility.
Original languageEnglish
JournalIEEE ISMAR
Publication statusPublished - 2016

Fingerprint Dive into the research topics of 'Image-based Indoor Localisation and Navigation with wearable AR devices'. Together they form a unique fingerprint.

Cite this