Purchase this article with an account.
Melissa Beck, Maura Lohrenz, J. Greg Trafton, Marlin Gendron; The role of local and global clutter in visual search. Journal of Vision 2008;8(6):1071. doi: https://doi.org/10.1167/8.6.1071.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
In visual search, preattentive processes locate potential target regions. Selective attention is then deployed to these regions to determine if the target is present. Generally, as the number of distractors in a display increases, this process becomes less efficient. The current studies examined the role of global visual clutter in the ability to deploy attention to target regions containing relatively more or less local clutter. Participants searched aerial maps with varying degrees of clutter for a predetermined target. A 2×3×3 factorial design was employed. The three factors examined were local clutter (high and low), global clutter (high, medium, and low) and distractor set size (4, 8, and 16). A clutter algorithm analyzing two components, color density and saliency, was used to determine levels of local and global clutter. Color density is a measure of how tightly packed similarly colored pixels are in the image and saliency is calculated using the distance in color space between two adjacent colors. The target on each trial was either an inverted V or an inverted W. When the target was an inverted V the distractors were inverted Ws and vice versa and the number of distractors varied across trials. Participants were faster to detect targets in low local clutter regions and faster to detect targets in low global clutter maps. There was also a significant local/global clutter interaction driven by larger effects of local clutter as global clutter increased. In addition, there was no effect of distractor set size on search time. Therefore, the preattentive process of detecting regions that are likely to contain the target is less efficient as the amount of global clutter increases. Furthermore, in complex images where it is difficult to determine the number of distractors, measuring global and local clutter may be a better predictor of search efficiency.
This PDF is available to Subscribers Only