Ranking of manipulated images in a large set using Error Level - - PowerPoint PPT Presentation

ranking of manipulated images in a large set using error
SMART_READER_LITE
LIVE PREVIEW

Ranking of manipulated images in a large set using Error Level - - PowerPoint PPT Presentation

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis Daan Wagenaar & Jeffrey Bosma University of Amsterdam In cooperation with the Netherlands Forensic Institute Ranking of manipulated images in a large set


slide-1
SLIDE 1

Ranking of manipulated images in a large set using Error Level Analysis

Daan Wagenaar & Jeffrey Bosma University of Amsterdam In cooperation with the Netherlands Forensic Institute

12-02-12

slide-2
SLIDE 2

Agenda

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

2

¤ Image Manipulation ¤ Research Question ¤ Error Level Analysis ¤ Methodology ¤ Experiments ¤ Results ¤ Conclusion ¤ Further Research ¤ Questions

slide-3
SLIDE 3

Agenda

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

3

¤ Image Manipulation ¤ Research Question ¤ Error Level Analysis ¤ Methodology ¤ Experiments ¤ Results ¤ Conclusion ¤ Further Research ¤ Questions

slide-4
SLIDE 4

Image Manipulation

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

4

¤ Examples

¤ Red Eye removal ¤ Brightness enhancements ¤ Sharpening ¤ …

¤ Most interesting manipulations

¤ Internal copy & move ¤ External copy & move

slide-5
SLIDE 5

5

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

Stalin with Yezhov (original) Stalin without Yezhov (manipulated) Object removal

slide-6
SLIDE 6

6

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

Object appearance modification Katie Couric (original) Slimmed Body (manipulated)

slide-7
SLIDE 7

7

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

Object addition Holding an iPhone (original) Holding a BlackBerry (manipulated)

slide-8
SLIDE 8

Agenda

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

8

¤ Image Manipulation ¤ Research Question ¤ Error Level Analysis ¤ Methodology ¤ Experiments ¤ Results ¤ Conclusion ¤ Further Research ¤ Questions

slide-9
SLIDE 9

Research Question

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

9

¤ Problem:

¤ A set of images as part of evidence ¤ An expert manually inspects each image for manipulations ¤ Time consuming process in a large set of images

v Can the Error Level Analysis technique be used to rank a set of images according to potentially present image manipulation?

slide-10
SLIDE 10

Agenda

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

10

¤ Image Manipulation ¤ Research Question ¤ Error Level Analysis ¤ Methodology ¤ Experiments ¤ Results ¤ Conclusion ¤ Further Research ¤ Questions

slide-11
SLIDE 11

Error Level Analysis (ELA)

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

11

¤ A technique for detecting image manipulations ¤ Uses properties of lossy image format ¤ Compares error caused by compression to a certain quality level ¤ An example:

¤ Initial image at a quality level of 95% ¤ ELA resaves this image at a certain quality level (e.g. 95%) ¤ Compression introduces error ¤ Compare error of initial and resaved image ¤ Manipulated areas will have a different level of error ¤ Differences are visibly expressed by brightness in a third image

slide-12
SLIDE 12

12

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

Original image ELA @75% ELA @ 85% ELA @ 95%

slide-13
SLIDE 13

13

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

Manipulated image ELA @75% ELA @ 85% ELA @ 95%

slide-14
SLIDE 14

14

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

¤ Limitations

¤ False positives can be caused by: ¤ Sharp contrast, well-defined patterns ¤ Recoloring, such as brightening, pallet skew, ... ¤ False negatives can be caused by: ¤ Low resolutions ¤ Scaling ¤ Low quality ¤ Image scanning from other sources ¤ Extremely skilled artists

slide-15
SLIDE 15

Agenda

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

15

¤ Image Manipulation ¤ Research Question ¤ Error Level Analysis ¤ Methodology ¤ Experiments ¤ Results ¤ Conclusion ¤ Further Research ¤ Questions

slide-16
SLIDE 16

Methodology

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

16

¤ Method 1: Average RGB values per block ¤ Method 2: Block to block comparison ¤ Method 3: Colored pixels ratio ¤ Method 4: Highest luminance value of the brightest pixel ¤ Method 5: Average luminance value of the 64 brightest pixels ¤ Method 6: Average luminance value of the brightest block

slide-17
SLIDE 17

Agenda

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

17

¤ Image Manipulation ¤ Research Question ¤ Error Level Analysis ¤ Methodology ¤ Experiments ¤ Results ¤ Conclusion ¤ Further Research ¤ Questions

slide-18
SLIDE 18

Experiments

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

18

¤ Goal ¤ Proof of concept ¤ Dataset of 300 images

¤ 100 images with Canon PowerShot A630 ¤ 100 images with iPhone 4 ¤ 100 images with Samsung Digimax S500

¤ 30 manipulated images

slide-19
SLIDE 19

Agenda

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

19

¤ Image Manipulation ¤ Research Question ¤ Error Level Analysis ¤ Methodology ¤ Experiments ¤ Results ¤ Conclusion ¤ Further Research ¤ Questions

slide-20
SLIDE 20

Results

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

20 50 100 150 200 250 300

Rank

20 40 60 80 100

Manipulated images found (%)

Method 3 (Colored pixels ratio) Method 4 (Highest luminance value of the brightest pixel) Method 5 (Average luminance value of the 64 brightest pixels) Method 6 (Average luminance value of the brightest block)

Rankings with ELA at 75%

slide-21
SLIDE 21

21

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

50 100 150 200 250 300

Rank

20 40 60 80 100

Manipulated images found (%)

Method 3 (Colored pixels ratio) Method 4 (Highest luminance value of the brightest pixel) Method 5 (Average luminance value of the 64 brightest pixels) Method 6 (Average luminance value of the brightest block)

Rankings with ELA at 85%

slide-22
SLIDE 22

22

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

50 100 150 200 250 300

Rank

20 40 60 80 100

Manipulated images found (%)

Method 3 (Colored pixels ratio) Method 4 (Highest luminance value of the brightest pixel) Method 5 (Average luminance value of the 64 brightest pixels) Method 6 (Average luminance value of the brightest block)

Rankings with ELA at 95%

slide-23
SLIDE 23

23

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

Manipulated image ELA @75% ELA @ 85% ELA @ 95%

slide-24
SLIDE 24

24

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

Manipulated image ELA @75% ELA @ 85% ELA @ 95%

slide-25
SLIDE 25

25

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

Manipulated image ELA @75% ELA @ 85% ELA @ 95%

slide-26
SLIDE 26

26

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

Original image ELA @75% ELA @ 85% ELA @ 95%

slide-27
SLIDE 27

27

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

ELA @75% ELA @ 85% ELA @ 95% Manipulated image

slide-28
SLIDE 28

Agenda

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

28

¤ Image Manipulation ¤ Research Question ¤ Error Level Analysis ¤ Methodology ¤ Experiments ¤ Results ¤ Conclusion ¤ Further Research ¤ Questions

slide-29
SLIDE 29

Conclusion

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

29

¤ Most effective method ¤ Limitations of ELA directly affect developed methods ¤ Detectable manipulation techniques v Can the Error Level Analysis technique be used to rank a set of images according to potentially present image manipulation?

¤ Yes, it is possible albeit not very reliable.

slide-30
SLIDE 30

Agenda

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

30

¤ Image Manipulation ¤ Research Question ¤ Error Level Analysis ¤ Methodology ¤ Experiments ¤ Results ¤ Conclusion ¤ Further Research ¤ Questions

slide-31
SLIDE 31

Further Research

12-02-12 Ranking of manipulated images in a large set using Error Level Analysis

31

¤ Alternative to ELA ¤ Combine different rankings ¤ Different methods

slide-32
SLIDE 32

Questions?

32

Ranking of manipulated images in a large set using Error Level Analysis 12-02-12