Volume IV-2/W5
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., IV-2/W5, 247-254, 2019
https://doi.org/10.5194/isprs-annals-IV-2-W5-247-2019
© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., IV-2/W5, 247-254, 2019
https://doi.org/10.5194/isprs-annals-IV-2-W5-247-2019
© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

  29 May 2019

29 May 2019

MODELLING UNCERTAINTY OF SINGLE IMAGE INDOOR LOCALISATION USING A 3D MODEL AND DEEP LEARNING

D. Acharya1, S. Singha Roy2, K. Khoshelham1, and S. Winter1 D. Acharya et al.
  • 1Department of Infrastructure Engineering, The University of Melbourne, Parkville, Victoria, Australia
  • 2Institute for Sustainable Industries and Livable Cities, Victoria University, Werribee, Victoria, Australia

Keywords: Indoor localisation, Camera pose regression, 3D models, Deep learning, Bayesian systems, Uncertainty

Abstract. Many current indoor localisation approaches need an initial location at the beginning of localisation. The existing visual approaches to indoor localisation perform a 3D reconstruction of the indoor spaces beforehand, for determining this initial location, which is challenging for large indoor spaces. In this research, we present a visual approach for indoor localisation that is eliminating the requirement of any image-based reconstruction of indoor spaces by using a 3D model. A deep Bayesian convolutional neural network is fine-tuned with synthetic images generated from a 3D model to estimate the camera pose of real images. The uncertainty of the estimated camera poses is modelled by sampling the outputs of the Bayesian network fine-tuned with synthetic images. The results of the experiments indicate that a localisation accuracy of 2 metres can be achieved using the proposed approach.