Academic journal article Human Factors

The Use of 2D and 3D Displays for Shape-Understanding versus Relative-Position Tasks

Academic journal article Human Factors

The Use of 2D and 3D Displays for Shape-Understanding versus Relative-Position Tasks

Article excerpt

Research on when and how to use three-dimensional (3D) perspective views on flat screens for operational tasks such as air traffic control is complex. We propose a functional distinction between tasks: those that require shape understanding versus those that require precise judgments of relative position. The distortions inherent in 3D displays hamper judging relative positions, whereas the integration of dimensions in 3D displays facilitates shape understanding. We confirmed these hypotheses with two initial experiments involving simple block shapes. The shape-understanding tasks were identification or mental rotation. The relative-position tasks were locating shadows and determining directions and distances between objects. We then extended the results to four experiments involving complex natural terrain. We compare our distinction with the integral/separable task distinction of Haskel and Wickens (1993). Applications for this research include displays for air traffic control, geoplots for military command and control, and potentially, any display of 3D information.


Many operational tasks require the comprehension of three-dimensional (3D) objects and environments. For example, air traffic controllers need to understand a complex 3D environment populated with aircraft, air routes, and no-fly zones. Similarly, military officers need to understand 3D environments such as ground and undersea terrain, air routes, and radar zones. 3D displays (or displays with 3D views) seem to provide a natural, and increasingly affordable, solution to these requirements.

What we mean by a 3D view is actually a perspective or oblique view of an object or scene displayed on a computer monitor. The image is two-dimensional (2D), but the viewing angle provides a 3D perspective. For example, rather than displaying an environment from directly above (a planar or bird's-eye view), perspective view technologies generally display the environment from a 30[degrees] or 45[degrees] angle. Holographic and other true 3D technologies are being developed, but most interest in 3D displays concerns 3D perspective views.

Many potential users who see 3D perspective views are enthusiastic about them. This appeal of 3D views most likely stems from their capability to convey the shape of complex objects in a natural and integrated way. Within the scientific visualization community, 3D perspective views have been used extensively to view complex objects such as molecules and even abstract objects such as the semantic structure of document collections (e.g., Card, Mackinlay, & Shneiderman, 1999).

However, Andre and Wickens (1995) caution system designers that sometimes "users want what's not best for them," preferring systems that hinder rather than enhance performance. Andre and Wickens' review of studies on input devices, display interfaces, color, and 3D displays provides evidence to support this cautionary note. Although it might be believed naively that more display dimensions are always better, there are three sources of problems with 3D views that stem from perspective projection.

First, without other depth cues available in the 3D view, the location of objects is ambiguous along lines of sight into the viewing plane. This problem has been termed projective ambiguity by Sedgwick (1986) and line-of-sight ambiguity by Boyer and Wickens (1994). Second, in addition to the linear affine transformations in a 2D view (magnification, translation, or both), space is nonlinearly distorted in a 3D view. Specifically, depth into the scene orthogonal to the picture plane is scaled nonlinearly (approximately quadratically) by depth, whereas widths and heights parallel to the picture plane are scaled linearly by depth (see Gillam, 1995). This asymmetric compression of space in a 3D view results in the distortion of distances and angles. Third, the projection of objects tilted toward the line of sight in a 3D view is compressed. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.