Abstract
There are many kinds of puzzle games on the Internet. In this thesis, three new steganographic methods are proposed based on three popular puzzle games: jigsaw puzzle, jig swap puzzle and (n2-1)-puzzle. The first novel method is proposed to hide secret data in jigsaw puzzle images. First, a digital image is taken as input and divided into blocks. Then, a semicircle is drawn and attached to the right/ bottom side of each block. The secret data are embedded through the attached positions and orientations of the semicircles according to a stegokey. The resulting image looks like those jigsaw puzzle images appearing on many jigsaw puzzle Web sites. The second novel method is proposed to hide a secret message in a jig swap puzzle image. First, a color image is taken as input and divided into blocks. Second, each block is rearranged to a new position according to the secret data and a stegokey. The resulted image is a perfect jig swap puzzle. The original image is needed for extracting secret data. Under the assumption that the receiver and the sender share some common images, the receiver can extract the secret data from the jig swap puzzle image. The third novel method is proposed to hide secret message in a solvable (n2-1)-puzzle. The secret message is first converted into a permutation of numbers 1 to n2-1. Based on a parity test algorithm and a stegokey, an even permutation generator is provided to transform the obtained permutation into a solvable stego-(n2-1)-puzzle. A scenario is also presented for secret message transmission via an on-line puzzle game. Experimental results show that all of these three proposed methods are undetectable. Furthermore, the first and second methods are robust to lossy compression and format conversion, and the third method has no robustness problem for lossy compression, format conversion and geometric distortion.