As the demand for three-dimensional image displays keeps growing continuously, so does the requirement for the objective evaluation of user experience. Multi-plane volumetric display is a new technology developed to visualize complex information in the form of spatial images in the volume of the optical element [1]. Visual information is perceived through a series of saccadic eye movements that are interrupted by visual fixations [2]. Most of fixations land on the certain objects that draw our visual attention [3]. Therefore, the evaluation of eye movements can provide useful insights about the quality of user experience for the novel displays.
To assess how individuals react on the third dimension of image provided by the multi-plane architecture of optical element, participants performed a visual search task on the multi-plane volumetric display. The study included target-absent trials to elucidate how observing three-dimensional and flat images affects eye movements. Search elements were displayed at four field eccentricities (2, 4, 6, and 8°) in a random order. In target-present trials, the target was presented per one plane closer to the participant in comparison to other elements. The task was to determine which of the four circles was displayed closer to the observer. In each trial, the image was displayed for two seconds, and at the end of this time, participants had to submit the response about the location of the target or report whether the target was absent. The viewing distance to the screen was 65 cm.
The results have shown that the performance accuracy is the highest in the target-present trials. However, more errors have been made in the target-absent trials. Evaluation of eye movements has revealed that if the three-dimensional effect is noticed immediately, the gaze is more often directed to the location of the target in comparison to other elements in the scene. On the other hand, when the spatial effect is not observed or it is more difficult to be noticed, the distribution of eye movements changes considerably. Specifically, similar amount of time is devoted to viewing each element.
To sum up, the obtained results indicate that it is easier to distinguish the three-dimensional images than the flat images on the multi-plane volumetric display if the viewing time and head movements are limited. To facilitate the comparison of displayed information in depth, the elements should be close to each other, because otherwise the three-dimensional images presented on sequential planes might be viewed and perceived as the two-dimensional ones, as shown by the results at higher field eccentricities. Thus, the distribution of eye movement parameters clearly shows whether a person has noticed the three-dimensional image effect. Therefore, the method base on the evaluation of eye movements can be further developed for the reliable assessment of user experience for the three-dimensional visualization systems.