Evaluating label placement for augmented reality view management

Abstract

View management, a relatively new area of research inAugmented Reality (AR) applications, is about the spatial layout of 2D virtual annotations in the view plane. This paper represents the first study in an actual AR application of a specific view management task: evaluating the placement of 2D virtual labels that identify information about real counterparts. Here, we objectively evaluated four different placement algorithms, including a novel algorithm for placement based on identifying existing clusters. The evaluation included both a statistical analysis of traditional metrics (e.g. counting overlaps) and an empirical user study guided by principles from human cognition.The numerical analysis of the three real-time algorithms revealed that our new cluster-based method recorded the best average placement accuracy while requiring only relatively moderate computation time.Measures of objective readability from the user study demonstrated that in practice, human subjects were able to read labels fastest with the algorithms that most quickly prevented overlap, even if placement wasn't ideal. MotivationOne of the primary goals of Augmented Reality (AR) is to associate virtual information with their counterparts in the real world. While the virtual information is usually rendered in 3D, in some applications the virtual information may be confined to the 2D view plane. For example, virtual annotations rendered in a head-mounted display might be 2D labels identifying the names of nearby buildings. The decision of where and how to draw the annotations, to determine the spatial layout in the view plane, is the problem of view management [5]. This is a relatively new area of research in AR and Mixed Reality applications.This paper focuses on a specific problem in view management: the placement and evaluation of 2D labels that are associated with 3D counterparts. A basic problem with drawing 2D labels is that the labels may obscure important background objects or may overlap, making them impossible to read. Figure 1 shows an example from our test application, with the labels arranged in randomly-chosen positions. This is a poor configuration because many labels overlap and are unreadable. Figure 2 shows a better labeling for exactly the same situation, where the labels were automatically repositioned by a cluster-based method described in Section 5. Figure 1: Initial (randomly chosen) label p o s i t i o n s Figure 2: Repositioned labels through clusterbased method (red dial pattern #2)Label placement for AR applications is not trivial. Even for a static image, the general label placement problem is NP-hard [7]. The number of possible label positions grows exponentially with the number of items to be labeled. For example, if a label can occupy one of 36 positions around an object, then for 20 objects there are 36 20 possible label positions (over 1x10 31 ). At interactive rates, only a few possible combinations can be considered. Furthermore, cognitive and perceptual issues regarding label placement for AR applications ar...

Automated positioning of 3D virtual scene in AR-based assembly and disassembly guiding system

“…Therefore, the layout of virtual annotations was one of the key techniques in this kind of AR applications. This technique was also called view management, which is about the spatial layout of 2D virtual annotations in the view plane [15]. Some layout methods of virtual annotations were presented to overcome overlap or inconsistency problems among annotations on the screen, such as the greedy depth first placement method, discrete gradient descent method, simulated annealing method, and cluster-based method [15].…”

Tangible Interfaces for Augmented Engineering Data Management

“…Several studies (Azuma et al, 2003;Götzelmann et al, 2007) have been presented in literature but only a few for technical annotations. This issue is even more critical in AR because the user's point of view and the background can vary continuously.…”

Natural Interaction Framework for Navigation Systems on Mobile Devices

“…Bell et al defined view management for interactive 3D user interfaces as of maintaining visual constraints on the projections of objects on the view plane, such as locating related objects near each other, or preventing objects from occluding each other (Bell et al, 2001). Azuma and Furmanski handled this discussion from 2D point of view and according to their research; view management is about the spatial layout of 2D virtual annotations in the view plane of augmented and mixed reality applications (Azuma & Furmanski, 2003). Other than the layout of annotations, text readability is affected from the interference of the background texture in the dynamically changing AR environment.…”