Spotlight on Off-Screen Points of Interest in Handheld Augmented Reality

Abstract

Navigating Augmented Reality (AR) environments with a handheld device often requires users to access digital contents (i.e. Points of Interests -POIs) associated with physical objects outside the field of view of the device's camera. Halo3D is a technique that displays the location of off-screen POIs as halos (arcs) along the edges of the screen. Halo3D reduces clutter by aggregating POIs but has not been evaluated. The results of a first experiment show that an enhanced version of Halo3D was 18% faster than the focus+context technique AroundPlot* for pointing at a POI, and perceived as 34% less intrusive than the arrowbased technique Arrow2D. The results of a second experiment in more realistic settings reveal that two variants of Halo3D that show the spatial distribution of POIs in clusters (1) enable an effective understanding of the off-screen environment and (2) require less effort than AroundPlot* to find POIs in the environment.

Arrow, Bézier Curve, or Halos? – Comparing 3D Out-of-View Object Visualization Techniques for Handheld Augmented Reality

“…Visualization techniques for displaying large 2D information spaces on 2D displays are typically categorized into overview+detail views, focus+context views, and contextual views [11,24,31]. While overview+detail techniques use a second view (e.g., a miniature map [13]) to give an overview of the complete information landscape, focus+context techniques and contextual views visualize the information relative to the users' view.…”

“…Tasks in handheld AR often require users to interact with points of interest (POIs) in close vicinity [31], such as searching for AR content in an interactive exhibition or calling a collaborator's attention to a specific object. However, due to the device cameras' limited field of view (FOV), handheld AR displays can be compared to movable peepholes that only show a small portion of the AR content.…”

“…However, due to the device cameras' limited field of view (FOV), handheld AR displays can be compared to movable peepholes that only show a small portion of the AR content. Consequently, the POIs are often off-screen, and users need to carefully scan their surroundings to find them [31]. Various out-of-view object visualizations have been introduced (e. g., [21,30,44]) to facilitate this process.…”

Exploration de la Physicalité des Widgets pour l’Interaction Basée sur des mouvements de la Tête le Cas des Menus en Réalité Mixte

“…• Les menus physiques reposent sur des objets physiques, dont l'ajout peut ne pas convenir à tous les contextes. Par exemple, ils ne sont pas adaptés aux environnements encombrés tels que les salles des machines ou les véhicules [19] et impliquent un environnement prédéfini n'autorisant pas l'interaction mobile [38]. • Même lorsqu'ils conviennent, les menus physiques ont un impact sur la spatialité de l'interface utilisateur car ils reposent sur des positions fixes spécifiques.…”

Target Expansion in Context

“…However, by projecting all the points, some Voronoi cells may be empty: these cells correspond to POIs not visible on screen that are located behind the camera, and whose projection are on screen. Coupled with an off-screen target localization technique like Halo3D [17], computing the Voronoi diagram for all the POIs in the environment could help users locate and select off-screen POIs. Focusing on the selection of POIs on screen as a first step, we calculate the Voronoi diagram only for POIs in front of users that are visible on screen inside the camera's field of view.…”