A New Steganographic Method for PaletteBased Images

1
A New Steganographic Method for Palette-Based
Images

• Present by ???

2
Outline

• Palette-based images
• Steganography using palette-based images
• Steganographic methods
• EzStego method and problems
• New steganographic method
• Experimental results
• Summary

3
Palette-Based Images

• A palette is a list of the RGB colors used in an
  image
• Most GIF files on the web today use an 8-bit
  palette
• This means that of all the 16,777,216 possible
  colors, there are only 256 RGB colors in the
  image
• The bit that is furthest to the right in a binary
  number is called the least significant bit (LSB)

4
Steganography Using Palette-Based Images

• Two approaches to message hiding in palette-based
  images
• Embedding messages into the palette
• Easy to design a secure method under some
  assumptions about the noise properties of the
  image source (a scanner, a CCD camera, etc.)
• The capacity is limited by the palette size
• Embedding into the image data
• Have higher capacity
• Hard to design a secure scheme

5
Steganographic Methods

• Permuting the image palette rather than changing
  the colors in the image
• Randomized palette will raise suspicion
• Resaving it may erase the information
• Limited capacity
• Hiding encrypted messages in the LSB of the
  palette colors (better approach)
• One of the most popular message hiding schemes is
  EzStego

6
EzStego Method (1/4)

• 1. Sort the palette
• It rearranges the copy of the palette so that
  colors that are near to each other in the color
  model are near to each other in in the palette

7
EzStego Method (2/4)

• 2. One bit at a time
• Find the index of the pixel's RGB color in the
  sorted palette
• Get one bit from the input file. Replace the
  least significant bit of the index
• Find the new RGB color that the index now points
  to in the sorted palette
• Find the index of the new RGB color in the
  original palette
• Change the pixel to the index of the new RGB color

8
EzStego Method (3/4)

• Example
• 17,231,31 is color 00100101 in the sorted
  palette
• The index value of 00100101 is changed to
  00100100
• Color 00100100 in the sorted palette is
  179,233,36
• 179,233,36 is color 11101110 in the original
  palette
• The pixel value is changed to 11101110

9
EzStego Method (4/4)

• 3. Recovering the input file
• Find the index of the pixel's RGB color in the
  sorted palette
• The least significant bit of the index came from
  the input file. Write it to the output file.
• Example
• 179,233,36 is color 00100100 in the sorted
  palette
• The least significant bit is 0
• Write 0 to the output file

10
EzStego Problems

• Colors with similar luminance values may be
  relatively far from each other
• Colors 6,98,233 and 233,6,98 have the same
  luminance but represent two completely different
  colors

11
New Steganographic Method (1/4)

• 1. The secret message m is first converted into a
  binary stream of length M

12
New Steganographic Method (2/4)

• 2. The M pixels for message embedding are chosen
  randomly using a pseudo-random number generator
  (PRNG) seeded with a secret key

13
New Steganographic Method (3/4)

• 3. The palette is searched for closest colors (in
  Euclidean norm) with the same parity as the
  message bit
• The distance between colors (R1G1B1) and (R2G2B2)
  is
• The parity of a color is defined as RGB mod 2

14
New Steganographic Method (4/4)

• 4. Extracting the secret message
• M pixels are selected using a PRNG seeded with
  user-defined seed
• The secret message is simply read by extracting
  the parity bits of the colors of selected pixels

15
The Embedding Example

• The first message bit is 0
• 17,231,31 is color 00100101 selected in the
  palette
• The first closet color is the same pixel
  17,231,31 and the parity is 1 (not matched)
• The second closet color is the pixel 179,233,36
  (the color 00100100 in the palette) and the
  parity is 0 (matched)
• The pixel color 00100101 change to 00100100

16
The Extracting Example

• The first pixel color selected is 00100100
• The parity bit is 0
• The first message bit is 0

17
Test Images (1/2)
18
Test Images (2/2)
19
RMS Distance
20
The Maximal Change
21
Summary

• The new technique causes approximately 4 times
  less distortion to the carrier image than EzStego
• The maximal color change is 4-5 times smaller for
  the new technique than that of EzStego