Reversible Image Watermarking Using Interpolation Technique Source: IEEE Transcation on Information Forensics and Security, Vol. 5, No. 1, March 2010 Authors: Lixin Luo, Zhenyong Chen, Ming Chen, Xiao Zeng and Zhang Xiong Speaker: Hon- Hang Chang
Reversible Image Watermarking Using Interpolation Technique
Feb 23, 2016
Reversible Image Watermarking Using Interpolation Technique . Source: IEEE Transcation on Information Forensics and Security, Vol. 5, No. 1, March 2010 Authors: Lixin Luo , Zhenyong Chen, Ming Chen, Xiao Zeng and Zhang Xiong Speaker: Hon- Hang Chang - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Reversible Image Watermarking Using Interpolation Technique
Source: IEEE Transcation on Information Forensics and Security, Vol. 5, No. 1, March 2010Authors: Lixin Luo, Zhenyong Chen, Ming Chen, Xiao Zeng and Zhang XiongSpeaker: Hon- Hang ChangDate: 2010. 10. 27
2
Outline
IntroductionProposed MethodExperiment ResultsConclusions
Introduction
3
Embed
Cover image
Watermark
Watermarked image
Watermark
Cover image
Watermarked image
Extract
4
Proposed Method(Cont.)
LSB replacement
LM, LNRM, RN
Boundary Map
Overhead
LSB replacement of the overhead information
Marginal area of cover-image
Cover image
5
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
Proposed Method(Con.t) Interpolation in Non-Sample pixels● Sample pixel
○ Non-Sample pixel
● ○ ● ○ ● ○ ●○ ○ ○ ○ ○ ○ ○● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○● ○ ● ○ ● ○ ●○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●Cover image X
○ The Non-Sample pixel after predicting
1-Level
2-Level
6
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
Proposed Method(Cont.) Interpolation in Sample pixels● Sample pixel
○ Non-Sample pixel
○ The Non-Sample pixel after predicting
● ○ ● ○ ● ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ○ ○ ○ ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ○ ○ ○ ○ ●
○ ○ ○ ○ ○ ○ ○
● ○ ● ○ ● ○ ●
3-Level
7
Proposed Method(Cont.) Interpolation in Non-Sample pixels (1/2)
60 91 30 20 20
24 52 18 45 60
50 47 40 43 50
75 45 67 22 50
50 60 60 40 30
S45= {60, 52,40}
Cover image X
Mean45=(S45 (1)+S45 (3))/2 =(60+40)/2 =50
Mean135=(S135 (1)+S135 (3))/2 =(30+50)/2 =40
S135={30, 52,50}
60 91 30 20 20
24 60
50 40 50
75 50
50 60 60 40 30
Interpolation X ’
7.99
))((31)( 2
3
14545
k
ukSe
3.83
))((31)( 2
3
1135135
k
ukSe
45 405448.0504552.0
40)()(
)(50)()(
)( 13545
45
13545
135
1351354545'
eee
eee
MeanwMeanwX
45 35
50 45
u= ( Mean45+ Mean135 )/ 2 = (50+40)/2 = 45
8
Proposed Method Interpolation in Non-Sample pixels (2/2)
60 91 30 20 20
24 52 18 45 60
50 47 40 43 50
75 45 67 22 50
50 60 60 40 30
S0= {30, 18,40}
Cover image X
S90={52, 18,45}60 91 30 20 20
24 60
50 40 50
75 50
50 60 60 40 30
Interpolation X ’
Mean0=(S0 (1)+S0(3))/2 =(30+40)/2 =35
Mean90=(S90 (1)+S90 (3))/2 =(52+45)/2 =48.5
0625.269
))((31)( 2
3
100
k
ukSe
5625.266
))((31)( 2
3
19090
k
ukSe
u= ( Mean0+ Mean90 )/ 2 = (35+48.5)/2 = 41.75
45 35
50 45
4246
4943
42 5.485429.0354571.0
48.5)()(
)(35
)()()(
900
0
900
90
909000'
××
eee
eee
MeanwMeanwX
9
Proposed Method Interpolation in Sample pixels
60 91 30 20 20
24 52 18 45 60
50 47 40 43 50
75 45 67 22 50
50 60 60 40 30
S0= {18, 40, 67}
Cover image X
S90={47, 40, 43}60 91 30 20 20
24 60
50 50
75 50
50 60 60 40 30
Interpolation X ’
Mean0=(S0 (1)+S0(3))/2 =(18+67)/2 =42.5
Mean90=(S90 (1)+S90 (3))/2 =(47+43)/2 =45
8958.405
))((31)( 2
3
100
k
ukSe
3958.8
))((31)( 2
3
19090
k
ukSe
45 459807.05.420193.0
45)()(
)(5.42)()(
)( 900
0
900
90
909000'
eee
eee
MeanwMeanwX
u= ( Mean45+ Mean135 )/ 2 = (42.5+45)/2 = 43.75
45 35
50 45
4246
494345
10
-3 -2 -1 0 1 2 3 402468
-3 -2 -1 0 1 2 3 402468
Proposed Method(Cont.) Embedding(Non-Sample pixels) (1/2)
Cover image X Interpolation X ’
1or 1' ,1
or ' ,0RMLMe
RMLMeb
RMLM
RM+1LN
Difference E
60 91 30 20 20 12 41
24 47 43 35 60 14 17
50 47 40 42 50 18 52
75 50 50 45 50 50 52
50 60 60 40 30 54 73
24 47 43 35 60 14 17
60 65 75 72 75 81 64
60 91 30 20 20 12 41
24 47 42 34 59 14 17
50 47 40 43 50 20 52
75 50 49 46 49 51 52
50 60 60 39 30 52 73
24 47 44 34 59 11 17
60 65 75 72 75 81 64
0 1 -1 0 0
0 -1 -2
0 1 -1 1 -1
0 1 2
0 -1 1 1 3
RN LM-1
LMRM
- =
11
60 91 30 20 20 12 41
24 47 42 34 59 14 17
50 47 40 43 50 20 52
75 50 49 46 49 51 52
50 60 60 39 30 52 73
24 47 44 34 59 11 17
60 65 75 72 75 81 64
60 91 30 20 20 12 41
24 46 43 33 58 13 17
50 47 40 42 50 18 52
75 49 51 45 51 50 52
50 60 60 40 30 54 73
24 46 43 35 61 14 17
60 65 75 72 75 81 64
Proposed Method(Cont.) Embedding(Non-Sample pixels) (2/2)
Interpolation X ’
1or 1' ,1
or ' ,0RMLMe
RMLMeb
Difference E
60 91 30 20 20 12 41
24 47 42 34 59 14 17
50 47 40 43 50 20 52
75 50 49 46 49 51 52
50 60 60 39 30 52 73
24 47 44 34 59 11 17
60 65 75 72 75 81 64
0 1 -1 0 0
0 -1 -2
0 1 -1 1 -1
0 1 2
0 -1 1 1 3
-3 -2 -1 0 1 2 3 402468 RMLM
RM+1LM-1
Difference E’
-1 1 -1 -1 -1
0 -1 -2
-1 2 -1 2 -1
0 1 2
-1 -1 1 2 3
W= 1 0 1 1 0 1 1 1 0 0 1 0 1
+
=
Interpolation X ’
Watermarked image
12
Proposed Method(Cont.) Embedding(Sample pixels)
60 91 30 20 20 12 41
24 46 43 33 58 13 17
50 47 40 42 50 18 52
75 49 51 45 51 50 52
50 60 60 40 30 54 73
24 46 43 35 61 14 17
60 65 75 72 75 81 64
Watermarked image
60 91 30 20 20 12 41
24 46 43 33 58 13 17
50 47 40 42 50 18 52
75 49 51 45 51 50 52
50 60 61 40 31 54 73
24 46 43 35 61 14 17
60 65 75 72 75 81 64
Interpolation X ’
0 0
-1 -1
Difference E
-2 -1 0 10
0.51
1.52
2.5
LN RN
LM RM
-2 -1 0 10
0.51
1.52
2.5LM RM
LM-1 RM+1
1or 1' ,1
or ' ,0RMLMe
RMLMeb
- =
13
Proposed Method(Cont.) Embedding(Sample pixels)
60 91 30 20 20 12 41
24 46 43 33 58 13 17
50 47 40 42 50 18 52
75 49 51 45 51 50 52
50 60 61 40 31 54 73
24 46 43 35 61 14 17
60 65 75 72 75 81 64
Interpolation X ’
0 0
-1 -1
Difference E
-2 -1 0 10
0.51
1.52
2.5LM RM
LM-1 RM+1
1or 1' ,1
or ' ,0RMLMe
RMLMeb
W= 1 0 1 1
1 0
-2 -2
Difference E’
60 91 30 20 20 12 41
24 46 43 33 58 13 17
50 47 40 42 50 18 52
75 49 51 45 51 50 52
50 60 61 40 31 54 73
24 46 43 35 61 14 17
60 65 75 72 75 81 64
Interpolation X ’
60 91 30 20 20 12 41
24 46 43 33 58 13 17
50 47 41 42 50 18 52
75 49 51 45 51 50 52
50 60 59 40 29 54 73
24 46 43 35 61 14 17
60 65 75 72 75 81 64
Watermarked image
+
=
14
Proposed Method(Cont.) Extracting(Sample pixels)
60 91 30 20 20 12 41
24 46 43 33 58 13 17
50 47 41 42 50 18 52
75 49 51 45 51 50 52
50 60 59 40 29 54 73
24 46 43 35 61 14 17
60 65 75 72 75 81 64
Watermarked image
60 91 30 20 20 12 41
24 46 43 33 58 13 17
50 47 40 42 50 18 52
75 49 51 45 51 50 52
50 60 61 40 31 54 73
24 46 43 35 61 14 17
60 65 75 72 75 81 64
Interpolation X ’
1 0
-2 -2
Difference E’
+
=
LM=-1RM=0
LN=-2RN=1
1or 1' ,1
or ' ,0RMLMe
RMLMeb
W2=1 0 1 1
0 0
-1 -1
Difference E
60 91 30 20 20 12 41
24 46 43 33 58 13 17
50 47 40 42 50 18 52
75 49 51 45 51 50 52
50 60 60 40 30 54 73
24 46 43 35 61 14 17
60 65 75 72 75 81 64
-
=
15
Proposed Method(Cont.) Extracting(Non-Sample pixels)
60 91 30 20 20 12 41
24 46 43 33 58 13 17
50 47 40 42 50 18 52
75 49 51 45 51 50 52
50 60 60 40 30 54 73
24 46 43 35 61 14 17
60 65 75 72 75 81 64
Watermarked images
60 91 30 20 20 12 41
24 47 42 34 59 14 17
50 47 40 43 50 20 52
75 50 49 46 49 51 52
50 60 60 39 30 52 73
24 47 44 34 59 11 17
60 65 75 72 75 81 64
Interpolation X ’
=-
Difference E’
-1 1 -1 -1 -1
0 -1 -2
-1 2 -1 2 -1
0 1 2
-1 -1 1 2 3
+
=
0 1 -1 0 0
0 -1 -2
0 1 -1 1 -1
0 1 2
0 -1 1 1 3
Difference E’
Cover Image X
60 91 30 20 20 12 41
24 47 43 35 60 14 17
50 47 40 42 50 18 52
75 50 50 45 50 50 52
50 60 60 40 30 54 73
24 47 43 35 60 14 17
60 65 75 72 75 81 64
LM=0RM=1
LN=-3RN=4
1or 1' ,1
or ' ,0RMLMe
RMLMeb
W1=1 0 1 1 0 1 1 1 0 0 1 0 1 W= W1 W∥ 2
16
To distinguish the Boundary pixel is corresponding to genuine or pseudo Pixel in cover image:
Proposed Method Boundary Map (B)
x=0
x’’ =-1 x’’ =1
x=255
x’’ =254 x’’ =256
X XUnderflow Overflow
Pixel in cover image:
Watermarked pixel:
B=…0To add ‘0’ in to the boundary mapx’’ =0 x’’ =255
x=1
x’’ =0 x’’ =2
x=254
x’’ =253 x’’ =255
B=…1To add ‘1’ in to the boundary map
Overflow and Underflow
17
Experiment ResultsTABLE I
COMPARISON RESULTS IN TERMS OF THE CAPACITY (bits) AND THE PSNR VALUE (dB) FOR LENA, BABOON, PLANE, AND SAILBOAT
18
Experiment Results
Fig. 1 Performance evaluation of multilayer embedding over standard in test image Lena
19
ConclusionsThe computation cost of the proposed method scheme is small.
The proposed scheme could guarantee high image quality without sacrificing embedding capacity.
Related Documents
