Author(s)

Bingyao Huang

Date Approved

12-3-2015

Embargo Period

12-16-2017

Document Type

Thesis

Degree Name

M.S. Electrical and Computer Engineering

Department

Electrical and Computer Engineering

College

Henry M. Rowan College of Engineering

Advisor

Ying Tang

Subject(s)

Augmented reality

Disciplines

Electrical and Computer Engineering

Abstract

Augmented reality (AR) is a technology that superimposes computer-generated 3D and/or 2D information on the user’s view of a surrounding environment in real-time, enhancing the user’s perception of the real world. Regardless of the field for which the application is applied, or its primary purpose in the scene, many AR pipelines share what might be thinned down to two specific goals, the first being range-finding the environment (whether this be in knowing a depth, precise 3D coordinates, or a camera pose estimation), and the second being registration and tracking of the 3D environment, such that an environment moving with respect to the camera can be followed. Both range-finding and tracking can be done using a black and white fiducial marker (i.e., marker-based AR) or some known parameters about the scene (i.e., markerless AR) in order to triangulate corresponding points. To meet users’ needs and demand, range-finding, registration and tracking must follow certain standards in terms of speed, flexibility, robustness and portability. In the past few decades, AR has been well studied and developed to be robust and fast enough for realtime applications. However, most of them are limited to certain environment or require a complicated offline training. With the advancement of mobile technology, users expect AR to be more flexible and portable that can be applied in any uncertain environment. Based on these remarks, this study focuses on markerless AR in mobile applications and proposes an AR system using one-shot structured light (SL). The markerless AR system is validated in terms of its real time performance and ease of use in unknown scenes.

Share

COinS