Visuospatial task performance as a function of two- and three-dimensional display presentation techniques

Three-dimensional (3D) displays are becoming more widely available and are being applied to a greater variety of human-computer interface domains. Pre vious research has shown that 3D display of objects and information often i s more appealing to the users, but for many tasks it is less useful than th e two-dimensional (2D) displays. New display techniques must be assessed fo r their ability to improve human operator performance. The purpose of this research was to compare human performance on several 2D and 3D display form ats across four visuospatial tasks. Qualified military and civilian air tra ffic controllers completed altitude and speed judgement tasks, a vectoring task, and a collision avoidance task on 2D top-down (plan-view), 3D perspec tive, 3D stereo, and laser-based 3D volumetric display systems. Each subjec t's speed and accuracy were measured on each task. Results indicated that t he 2D plan or side-view displays yielded performance as good or better than any other display system for speed and altitude judgment tasks. Data prese ntation on the 3D volumetric display was superior to 3D perspective, 3D ste reoscopic, and 2D displays only for the collision avoidance task. These res ults support previous research suggesting that 3D displays are useful in ve ry specific tasks. The results from the collision avoidance experiment sugg est that tasks requiring operators to view and predict future locations of multiple display symbols traversing a confined space (such as relationships between aircraft within the airspace around an airport) appear to be well suited for 3D rendering. Compared to 3D stereoscopic and perspective displa ys, the veridical display of localized spatial information within a volumet ric display may provide high fidelity stereoscopic and parallax cues, impro ving human performance for some tasks. (C) Published by Elsevier Science B. V.