Purchase this article with an account.
Alyssa Hess, Mark Neider; Guiding search for camouflaged targets: Does color matter?. Journal of Vision 2014;14(10):922. doi: 10.1167/14.10.922.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Visual search in the real world often involves identifying a target that blends into its environment. Often, in such cases, a perfect target template is unavailable for guiding search. Previously, we characterized search for camouflaged targets in natural scenes with and without a target preview (Hess et al., 2013), and found little evidence for a preview benefit. Previous research indicates that color is an important feature in guiding visual search. To determine whether this is the case in search for camouflaged targets, where targets are often poorly defined, we replicated our previous studies where participants searched for a camouflaged target in natural scenes, but in the current experiments search scenes were presented in grayscale. Manual response data indicated that, across four target sizes and experimental blocks, participants got faster to find the target with practice and with decreasing difficulty (manipulated via target size) in target present trials, regardless of whether or not they received a target preview (all p <.05), trends consistent with our previous findings in full color scenes. Accuracy data followed a similar pattern; however, no time-based improvements were found without preview. To characterize the influence of color information in guiding search, we compared the data from the current grayscale studies to those of our previous color studies. Regardless of whether a target preview was presented, there was no significant response time difference between color and grayscale scenes (all p > .39). However, when a target preview was provided, participants were more accurate in color compared to grayscale scenes (p <.001). Overall, our data suggest that during camouflage search, if color information regarding the target is provided observers utilize that information; in the absence of such information, however, target representations are based largely on alternative features, even as participants become more familiar with the target set.
Meeting abstract presented at VSS 2014
This PDF is available to Subscribers Only