Camouflage and visual perception
Tom Troscianko1, Christopher P. Benton1, P George Lovell1,
David J. Tolhurst2 and Zygmunt Pizlo3
Supplementary Electronic Materials:-
2. Edge detection processes: disruption and camouflage
Figure 3:
Figure 3: A and B, two examples of illusory shapes or illusory contours. In A, two high contrast geometric stimuli seem to cap the ends of a bar; an illusory bright bar can be sensed. The geometric figures of A can be seen to "match" the ends of a receptive field (see C) and can be presumed to stimulate the edge detector weakly, thus leading to the weak sensation of the presence of a bright bar. D, this illusory contour (caused by the abutting of hatching at different angles) does not seem to be an appropriate stimulus for such edge detectors. The dark stimuli in C and D are shown in lighter grey so that the receptive field structure can still be seen.
4 Objects and shape
Figure 4:
Figure 4. A is a 2D image of a mantis. B shows contours extracted by hand superimposed on the image. C and D show two images of the recovered 3D shape of the mantis. This example was prepared by Tadamasa Sawada and Yunfeng Li.
5 Visual search – features across the scene
Figure 6:
Figure 6, Duncan and Humphreys (1989,1992) predicted that search-slope (the amount that search time increases as a function of each additional distractor) varies as a function of target-distractor similarity and distractor-distractor similarity. In each of the four examples (inset) the target is the central horizontal bar - though obviously in an experimental setting the location of the target would have been randomised. The search-slope varies as a function of the properties of the surrounding bars.
Figure 7:
Figure 7, Estimates of target-stimulus difference were achieved by comparing the visual-difference between target (a) and the visual scene (b). The differences are calculated using a VDP model at each spatial-frequency and within each colour-opponent channel (see text).
- 1 -