Purchase this article with an account.
Henry L. Apfelbaum, Doris H. Apfelbaum, Russell L. Woods, Eli Peli; The effect of edge filtering on inattentional blindness. Journal of Vision 2005;5(8):547. doi: https://doi.org/10.1167/5.8.547.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Neisser & Becklen (1975, Cognitive Psychology 7, 480–494) identified inattentional blindness; the inability of observers to maintain awareness of events in more than one of two superimposed scenes. The ability of the brain to make use of such multiplexed visual information and avoid confusion is central to the utility of many augmented vision devices, such as see-through head-mounted displays and head up displays. We are developing several such devices to aid people with vision impairments. Visual or perceptual phenomena such as binocular or perceptual rivalry and inattentional blindness may limit the utility of these devices. Many aspects of the display format might possibly affect these phenomena. Specifically, if the two scenes are easier to separate perceptually they may not rival to the same degree.
In this study we investigated the effect of edge filtering on inattentional blindness and the ability to follow superimposed/multiplexed scenes. We closely reproduced parts of the original Neisser & Becklen experiment, and then treated one or both of the video scenes with edge filtering to create a cartoon-like image. The special bipolar edge filtering produced white and black contours at each luminance edge, facilitating clear uninterrupted visibility of the edge-filtered scene over bright and dark sections of the other scene.
Normally-sighted young adults (n=36) viewed overlaid videos that included 6 trials with unexpected events, while attending scenes that did not have the events. Edge filtering was applied to the attended scene in two of the trials and to the unattended scene in two trials. We found no evidence that edge filtering affected the detection of unexpected events. However, filtering the unattended scene improved performance of the attended task, as measured by response time to actions in the attended scene. Filtering the attended scene reduced performance of the attended task.
This PDF is available to Subscribers Only