A majority-voting based watermarking scheme for color image tamper detection and recovery - PowerPoint PPT Presentation

Loading...

PPT – A majority-voting based watermarking scheme for color image tamper detection and recovery PowerPoint presentation | free to download - id: 1f8367-ZDc1Z



Loading


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation
Title:

A majority-voting based watermarking scheme for color image tamper detection and recovery

Description:

A majority-voting based watermarking scheme for color image tamper ... Ming-Shi Wang *, Wei-Che Chen. 11 November 2006. present: ???. Outline. Introduction ... – PowerPoint PPT presentation

Number of Views:85
Avg rating:3.0/5.0
Slides: 18
Provided by: Supe1
Category:

less

Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: A majority-voting based watermarking scheme for color image tamper detection and recovery


1
A majority-voting based watermarking scheme for
color image tamper detection and recovery
  • Ming-Shi Wang , Wei-Che Chen
  • 11 November 2006
  • present ???

2
Outline
  • Introduction
  • Proposed scheme
  • Block-based watermark embedding
  • Watermark verification
  • Tampered image recovery
  • Experimental results

3
Introduction
  • Two categories of conveying the authentication
    data labeling-based schemes and watermarking
    based schemes.
  • The watermarking-based authentication scheme can
    be classified as either fragile watermarking,
    which is designed to detect any possible
    modification of the pixel values, or semi-fragile
    watermarking, which just distinguish
    content-preserving operations form malicious
    alteration.

4
Proposed scheme
5
Block division
MM
Zij
8-bit R
8-bit B
8-bit G
6
Example (Block mapping)
  • block Z1,1 k12, k23, M10
  • So the recovery information of Z1,1 will hide in
    block Z2,1

7
Block-based watermark embedding
  • In the proposed scheme, a
  • 36-bit watermark consists
  • of the authentication data( 16 bits) and the
    recovery data( 20 bits) is embedded into the LSBs
    of each image block.
  • The authentication data of length 16 bits is
    obtained by exploiting both the local features of
    each image block and the global features of the
    image.

8
Example (local feature)
the total number of 1s in the five MSBs of all
the color components of pixel l
The parity-check bit for pixel l
01010101
10110110
1
11110001
so
0101010110111101
The local feature of block Z1,1
0000000000000001
0000000000000001
9
Example (global feature)
  • Hash function, MD5

01101101101100110
converted from 128 to 16 bits
d0d1d2d16 0110110110110011
the global features of image Z
10
Authentication data
  • Authentication local (xor) global

11
Example (recovery data)
01010101
10110110
1
11110001
(u,v) is mapped to (i,j)
That is because we just want 20 Bits for recovery
data.
12
Example (recovery data)
  • Assume Yu,v01010011, Cbu,v00001110
  • Cru,v11011101 ,(u,v) mapped to(1,1)
  • So the recovery data in block Zi,j is
  • C1,101010011000011110111

13
Example(LSB embedding)
  • authentication data(16)
  • 0011010101010101
  • recovery data(20)
  • 00101101001101010011

R001 B101 G010
R101 B010 G100
R101 B101 G001
R101 B010 G011
14
Watermark verification and Tampered image
recovery
15
Experimental results(1/3)
  • PSNR37.96

PSNR37.96
PSNR37.83
16
Experimental results(2/3)
  • Rtd100

Rtd100
Rtd100
17
Experimental results(3/3)
About PowerShow.com