Adaptive Data Hiding in Palette Images by Color

Adaptive Data Hiding in Palette Images by Color Ordering and Mapping With Security Protection Author : Chih-Hsuan Tzeng, Zhi-Fang Yang, and Wen-Hsiang Tsai Source : IEEE Transactions on communications, Vol. 52, No. 5, May 2004, pp. 791 - 800 Speaker: Z. Y. Wu(吳紫嫣) Date : 2005/05/10

Outline l l l l Palette Image Replacement by the neighboring color to hiding data Color-Ordering Relationship Color-Mapping function Embeddable and non-embeddable pixels Embedding process and Extraction process Experimental Results Conclusions 2

Introduction l Palette Image: Color Palette (256 colors) 0 15 31 16 Pixel color mapping 32 47 48. . 63. . 240 255 Color Indexes C 8 C 8 C 9 . . . C 7 . . . 3

Replacement by the neighboring color to hiding data c 1: (0, 174, 239), c 2: (57, 181, 74), c 3: (236, 0, 140) 4

Color-Ordering Relationship l l l Two colors (RGB values): l c 1: (r 1, g 1, b 1) = (200, 100, 200) l c 2: (r 2, g 2, b 2) = ( 40, 200, 100) The luminance value v 1 and v 2 of c 1 and c 2: l v 1= 0. 3*r 1+0. 59*g 1+0. 11*b 1= 141 l v 2= 0. 3*r 2+0. 59*g 2+0. 11*b 2 = 141 The color-ordering relationship Rco: c 1 > c 2 (v 1=v 2 and r 1>r 2) 5

Color-Mapping function l The color-mapping function c 1 c 2 c 3 (150, 100, 50) (100, 200, 100) (100, 150, 100) v 1= 110; v 2= 159; v 3= 130; v 4= 100 (100, 100) (150, 100, 200) v = 126 c 4 c 6

Embeddable and non-embeddable pixels Non-embeddable (α=1< Tc) (100, 100, 100) (110, 90, 100) (100, 100) (120, 150, 60) (100, 100) X 1 X 2 Ex. Tc=2; Td=15 Non-embeddable β=67 > Td (100, 100) (105, 100, 98) (95, 105) (105, 100, 98) X 3 α : the number of distinct colors of four neighbors Embeddable(α=3; β=13 ) β : the maximum color difference between X and four neighbors. 7

Embedding process(1/3) l l l Step 1) For each secret bit bj in S , perform the following steps until all secret bits in S are embedded. Step 2) Perform a raster scan of I and check the data embeddability of each scanned pixel, until a data-embeddable pixel X is found. Step 3) Take the color c of X and the sorted colors c 1’-c 4’ of the four precedent neighbors of as input to the color-mapping function fcm to yield a binary output bit b 0. Step 4) Check whether the secret bit bj is equal to b 0. If so, regard the secret bit bj to be already existing at X, and go to Step 1) to embed the next bit; otherwise, perform the next step. Step 5) Find the optimal replacement color copfor X by Algorithm 1; substitute the color c of X with copt ; and go to Step 1) to embed the next bit. 8

Embedding process(2/3) S = b 0 b 1 b 2, …, bn (secret bit stream) (c 1, c 2, c 3, c 4) sorted colors Color-ordering relationship (c 1’, c 2’, c 3’, c 4’) Color-mapping function Optimal replacement color copt for X b 0= b j yes embedded no Ex. b 0 = Fcm(c, c 1’, c 2’, c 3’, c 4’)= 0 c c c 1 2 3 (150, 100, 50) (100, 200, 100) (100, 150, 100) (100, 100) (150, 100, 200) c 4 X bj= 0 (secret bit) 9

Embedding process(3/3) Ex. b 0 = Fcm(c, c 1’, c 2’, c 3’, c 4’) = 0 c c c 1 2 3 (100, 100) (105, 100, 98) (95, 105) (100, 105) (105, 100, 98) c 4 X bj= 1 (secret bit) bj b 0 find c’: ( 100, 105) 10

Extraction process (c 1, c 2, c 3, c 4) Color-ordering sorted colors relationship (c 1’, c 2’, c 3’, c 4’) bit) c 1 c 2 Color-mapping function b (secret c 3 (150, 100, 50) (100, 200, 100) (100, 150, 100) (100, 100) (150, 100, 200) c 4 X b= 0 (secret bit) 11

Experimental Results 12

Conclusions l l The major idea of the proposed data-embedding process is to modify the colors of data-embeddable image pixels so that the binary outputs of the color-mapping function with the colors of these image pixels as input may be taken as the data to be hidden. Using the color-mapping function based on color-ordering relationship. 13