Camouflage Images
Hung-Kuo Chu, Wei-Hsin Hsu, Niloy J. Mitra, Daniel Cohen-Or, Tien-Tsin Wong, Tong-Yee Lee


Camouflage images contain one or more hidden figures that remain imperceptible or unnoticed for a while. The ability to delay the perception of the hidden figures is attributed to the theory that human perception works in two main phases: feature search and conjunction search. Effective camouflage images make feature based recognition difficult, and thus force the recognition process to employ conjunction search, which takes considerable effort and time. In this paper, we present a technique for creating camouflage images. To fool the feature search, we remove the original subtle texture details of the hidden figures and replace them by that of the surrounding background. To leave an appropriate degree of clues for the conjunction search, we compute and assign new tones to regions in the embedded figures by performing an optimization between two conflicting terms, namely immersion and standout, corresponding to hiding and leaving clues, respectively. We show a large number of camouflage images generated by our technique, with or without user guidance. We have tested the quality of the images in an extensive user study, showing a good control of the difficulty levels.


(Top) Two camouflage images produced by our technique. The left and right images have seven and four camouflaged objects, respectively, at various levels of difficulty. By removing distinguishable elements from the camouflaged objects we make feature search difficult, forcing the viewers to use conjunction search, a serial and delayed procedure. (Please zoom in for a better effect.)
(Left) Artist created camouflage images: (left) 8 eagles, and (right) 13 wolves are embedded (Copyright of Steven Gardner)

Mouse over the image to see embedded objects.

(Below) Results of camouflaging a lion onto a mountain backdrop using various methods: (left to right) alpha blending, Poisson cloning, texture transfer, Poisson cloning followed by texture transfer, and our method.
(Below) Three camouflage images created by our algorithm. (Mouse over the image to see embedded objects.)

(Top) Recognition time and success rates on three difficulty levels of generated camouflage images as observed in course of our user study (see Section 6).
(Top) Comparison of synthesized results (right) with artist generated ones (left).


AUTHOR = "Hung-Kuo Chu and Wei-Hsin Hsu and Niloy J. Mitra and Daniel Cohen-Or and Tien-Tsin Wong and Tong-Yee Lee",
TITLE = "Camouflage Images",
JOURNAL = "ACM Transactions on Graphics",
VOLUME = "29", 
NUMBER = "3",
YEAR = "2010"
NOTE = "to appear"

paper (70MB) paper (7MB) slides (54MB)
back to publications
back to homepage