An Estimation Approach for Tracking Head Orientation while Navigating Street View on the Flat Screen Monitor

• Main Authors: Ping-Jung Lin, 林品鐘
• Other Authors: Wei-Chung Teng
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/82141435741949920675

碩士 === 國立臺灣科技大學 === 資訊工程系 === 103 === In this thesis, we adopted Google Maps as our platform to study the differences in eye movements as well as the performance of test subjects when asked to perform several tasks using a traditional flat screen monitor and using a widely-known head-mounted virtual reality display Oculus DK2. There are several reasons why comparing the performance between these two user interfaces is a difficult task. First, the street views that Google Map provides are panorama images attached to a spherical 3D model, and only a part of it is showed on a flat screen monitor. Therefore, simply recording the coordinates measured by eye tracking devices will not help us understand where the users are really looking at. Second, when wearing a head-mounted VR display such as Oculus DK2, there is no way to track eye movements as the eyes are blocked by the display itself. In order to overcome these difficulties, we developed a method to transform the data collected from these two user interfaces into one common format. Experiment results show significant differences between the two user interfaces. Our test subjects show 75.8% increment in fixation time, 37.0% decrement in saccade time and 17.3% decrement in time spent on checking maps while wearing Oculus DK2 which is better than using a traditional flat screen monitor.