Evaluation of Drop Shadows for Virtual Object Grasping in Augmented Reality
PubDate: May 2020
Teams: Birmingham City University
Writers: Maadh Al-Kalbani; Maite Frutos-Pascual; Ian Williams
This article presents the use of rendered visual cues as drop shadows and their impact on overall usability and accuracy of grasping interactions for monitor-based exocentric augmented reality (AR). We report on two conditions, grasping with drop shadows and without drop shadows, and analyze a total of 1620 grasps of two virtual object types (cubes and spheres). We report on the accuracy of one grasp type, the Medium Wrap grasp, against Grasp Aperture (GAp), Grasp Displacement (GDisp), completion time, and usability metrics from 30 participants. A comprehensive statistical analysis of the results is presented giving comparisons of the inclusion of drop shadows in AR grasping. Findings showed that the use of drop shadows increases usability of AR grasping while significantly decreasing task completion times. Furthermore, drop shadows also significantly improve user’s depth estimation of AR object position. However, this study also shows that using drop shadows does not improve user’s object size estimation, which remains as a problematic element in grasping AR interaction literature.