Abstract
For humans, many tasks necessitate the ability to detect camouflaged objects from their background, (e.g., hunters searching for camouflaged prey). Edge detection is critical in breaking the camouflage of an object that matches the color, luminance, and pattern of its background. We investigate how edge complexity, in the form of different shapes defined by radial frequency (RF = 3, 4, 5, and 8) and amplitude (0.1, 0.25, and 0.5), affects target ‘survivability’ in a computer based visual search task. The task in each trial was to rapidly select five randomly located, identically shaped targets within a 3-minute time limit. The targets and backgrounds were pink noise (1/f) textures. The presentation order of the different conditions (4 RF x 3 amplitude + 1 control circle, RF0) was randomized (one condition per trial) with two repeats each. In Experiment 1, the targets were surface-area-matched while in Experiment 2, they were angular-size-matched (i.e., visual angle across the widest point). Data were analyzed using survival analysis (targets that were harder to detect had greater survival). In general, target survival increased with increasing radial frequency when amplitude was low, but there was no clear trend when amplitude was high. Moreover, overall target survivability was greatest at high radial frequency with low amplitude, and worst at high radial frequency with high amplitude. These overall trends held across both experiments, indicating that the differences in survival between conditions were driven by radial frequency and amplitude and were not simply an effect of angular-size or surface area. These results demonstrate that the ability to detect a camouflaged object is affected by an interaction between its radial frequency and amplitude. We therefore expect these two factors to be subject to natural selection and consequently they may be important drivers in the evolution of camouflage strategies in nature.