Greylevel Image Hiding Scheme Based on Vector Quantization

Grey-level Image Hiding Scheme Based on Vector Quantization Author : Yu Chen Hu Source : Electronics Letters , Vol. 39 , 23 Jan. 2003, pp. 202 - 203 Date : 2005/04/26 1

Outline l Motivations l Proposed Scheme l Simulation Results l Conclusions 2

Motivations l Hiding capacity l Computational cost 3

Proposed Scheme 4

Encryption l l l w , h : width and height of the cover image and the secret images. t : number of secret images. Nc : size of codebook. k : length of codeword. d : selection parameter of initial codewords of LBG. r : the modified bits r of each cover pixel. 5

Encryption (cont. ) 1 0 0 1 1 1 0 0 Replace last r bits with 2(r-1) Original Image C C’ 6

Encryption (cont. ) C’ 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 S 1 31 32 33 34 35 36 VQ S 2 VQ d=3 LBG i 1 codebook i 2 i 1 i 2 I 7

Encryption (cont. ) Original Image C DES Hiding into the rightmost two bits of the pixels in the first row (w, h, t, Nc, k, d) DES Hiding into the rightmost r bits of the pixels in the (h-1) rows E I 8

Hiding Capacity 9

Original Image PSNR = 30. 70 Embedded image PSNR = 45. 53 d. B PSNR = 30. 39 PSNR = 32. 19 10

Conclusions l This scheme is safe because the DES cryptosystem is employed to protect the working parameters and indices. l Small codebook reduce the computational cost. 11