UBIQUITOUS COORDINATED AND MULTIPLE VIEWS

Abstract

This thesis explores the concept of Ubiquitous Coordinated and Multiple Views (UCMV), or the linking of related static and/or digital views that are distributed in a work environment, we introduce a prototype system that uses the medium of augmented reality (AR) on handheld devices as a means of drawing correlations across paper documents and retrieving related digital content. The thesis focuses on identifying the techniques suitable for selecting items or regions from a document or view. We consider three selection techniques: Device Pointing, Touch on Screen and Finger Pointing. With the Device Pointing technique, the mobile device is itself used as a pointer to make a selection by tapping directly on the document. The Touch on Screen technique uses the mobile device’s screen to make selections. Finally, the user points their finger at a desired document region to make a selection in the Finger Pointing technique. We present early design work and preliminary prototype evaluations leading to a set of expected strengths and weaknesses for each of the three techniques. We hypothesized that no one technique would be ideal for the full range of selection operations expected. A formal study was then conducted in order to provide evidence for our hypothesis, grounded in an airplane mechanic scenario. Results show that Device Pointing was the fastest and most preferred technique for discrete part selection while the Touch on Screen technique was the fastest and most preferred for region selection on large documents. We discuss implications for future UCMV designs, and consider limitations of our study.