Loading
Research Projects
SwingAR: An Augmented Reality Approach for Indoor-Object Positioning
|
Abstract
The unavailability of an accurate GPS positioning technology in indoor environment has paved way for research with the goal of providing location-based services or object positioning services. There have been many attempts to exploit the various on-board sensors of a smartphone and create opportunities for providing location-based services without using GPS. However, these techniques either need a priori information such as floor-plan or rely on war-driving and incur a bootstrapping dull period. In this proposal we propose a system which does not assume anything about the indoor environment and that can provide a very accurate object localization. Specifically, this proposal intends to address the following problem: can we localize an object by looking at an object through the smartphone camera? As an application, we build an augmented reality application using which a user can associate a tag to an object and retrieve the tags by looking through her smartphone's camera. Keeping smartphone battery limitation in mind, our core-idea to localize objects does not use any computer vision algorithm.
Advisor(s): Landon Cox
Romit Roy Choudhury, Bruce Maggs