Thesis (Ph. D.)--University of Hawaii at Manoa, 1994. / Includes bibliographical references (leaves 273-281). / Microfiche. / xv, 281 leaves, bound ill. 29 cm

Recent applications in three-dimensional (3-D) presentation of information has emphasized the use of field-sequential stereoscopic CRT displays. The usefulness of stereopsis to present depth information has been studied in terms of performance measures such as search time, search accuracy, and subjective image quality ratings. Few, if any, studies have been conducted to investigate the influence of stereopsis upon cognitive workload. This dissertation is a description of two experiments using 3-D images presented on a Tektronix SGS 620 field-sequential stereoscopic CRT. Comparisons were made between presentations incorporating monocular cues only (2 1/2-D) and scenes containing monocular cues plus retinal disparity (3-D). In the first experiment, 11 participants were required to make inter-elements distance judgments under conditions of varying perspective angle, scene complexity, and depth cues while performing an interval production task. Accuracy was found to be significantly enhanced by the addition of retinal disparity as well as by decreases in the complexity of the presented scene. Cognitive workload increased significantly with increases in scene complexity, but no significant difference in workload was found between the stereoscopic and non-stereoscopic presentation formats. The second experiment investigated the effects of stereopsis on target identification under varying conditions of background complexity, target coding schemes, number of elements, signal-to-noise ratio, and depth cues. Accuracy measures, response time, and workload ratings were degraded by increases in scene background complexity, signal:noise ratio, and number of elements. Target coding methods which incorporated color were found to be significantly better than those in which targets were identified by shape only. No significant difference was found between 2 1/2-D and 3-D presentation for any of the dependent measures. The results indicate that retinal disparity it most appropriate for tasks, such as determination of object location in depth, in which the additional information provided by the depth cue is directly related to the task. Retinal disparity provides the most benefit for depth judgment tasks involving highly complex scenes. In cases of very simple tasks or simple scenes, the addition of retinal disparity is contra-indicated. / Ph. D.