Image Forensics of High Dynamic Range Imaging 10th International - - PowerPoint PPT Presentation

Image Forensics of High Dynamic Range Imaging

10th International Workshop on Digital-Forensics & Watermarking

This research is sponsored by an EPSRC/Charteris CASE Award

• P. J. Bateman, A. T. S. Ho, and J. A. Briffa

Image Forensics

Uncovering facts about an image without actively injecting data Verify the integrity of a digital image

Source Classification Camera Identification Processing History Recovery Forgery Detection Anomaly Investigation

• M. Chen, J. Fridrich, M. Goljan, and J. Lukás, “Image Origin and Integrity Using Sensor Noise,” IEEE Transactions on Information Security and

Forensics, 3(1), pp. 74-90, March 2008

• H. Farid, “Digital Image Forensics”, American Academy of Forensic Sciences, Washington, DC, 2008

Auto-bracketing Camera Merging and Registration Tone Mapping LDR version of HDR Image

HDR Pipeline

HDR

• 1EV

0EV +1EV

• E. Reinhard, G. Ward, S. Pattanaik, and P. Debevec, “High Dynamic Range Imaging: Acquisition, Display, and Image-Based Lighting,” Morgan Kauffman,

ISBN: 978-0-12-585263-0, 2005.

HDR Imaging

High Dynamic Range Imaging

High Dynamic Range Imaging

High Dynamic Range Imaging

High Dynamic Range Imaging

Image Histogram LDR HDR

30,000 60,000 255 350,000 700,000 255

High Dynamic Range Imaging

HDR is Popular

Stats taken from Flickr.com (24-October-2011)

iPhone 4 Camera Useage

Can we detect HDR- Processed Images?

• An increasingly popular photography method
• On-board implementations
• More and more HDR images will exist amongst LDR
• EXIF metadata shows little regarding HDR processed images.
• The HDR pipeline differs from manufacturer to manufacturer
• Do images contain fingerprints of specific manufacturing pipelines?
• Images can look heavily processed, but are straight off camera
• This may fool existing Forensic algorithms
• A novel subject of Image Forensics

Research Motivation

Processing History Recovery Anomaly Investigation Camera Identification

HDR Detection in Image Forensics

Auto-bracketing Camera Merging and Registration LDR version of HDR Image Tone Mapping

HDR Pipeline

• Operates on Illuminance-Reflectance model
• illuminance can be reduced
• Separate illuminance and reflectance components

I(x,y) = i(x,y) · r(x,y) D = log i(x,y) + log r(x,y)

• (High-Pass) filtering in FFT domain is applied*
• Attenuate low frequencies (illuminance)
• Preserve high frequencies (reflectance)

*A. V. Oppenheim, R. Schafer, and T. Stockham, “Nonlinear Filtering of Multiplied and Convolved Signals,” in Proceedings of the IEEE, 56(8), pp. 12641291, 1968.

Homomorphic Filtering

“Strong” edges contain high and low frequency data Haloing artefacts are produced

• G. Qiu, J. Guan, J. Duan, M. Chen, “Tone Mapping for HDR Image using Optimization: A New Closed Form Solution,” 18th International Conference on

Pattern Recognition, pp. 996-999, 2006.

The Problem with HF

Aim: To accurately classify HDR/LDR images Device Used: Apple iPhone 4 (Native Camera App) Method: Capture 100 real-world “landscape” images

• 50 HDR
• 50 LDR

Images are captured from a tripod to ensure registration processing is minimised

The Experiment

HDR Image

LDR Image

LDR HDR

Spatial Pixel Distribution

The Strategy

Read Image

(extract luminance)

Canny Edge

Remove Texture Find “Strongest” Edge

FFT Edge Data

Classify Edge Data Majority Voting Output

• 1. Read Image and Extract

Luminance

• 2. Canny Edge Detection

• 3. Threshold Y to B&W

• 4. Morphology: “Open”

• 5. Sobel Edge Detection

• 6. Morphology: “Open”

Remove connected edges that do not satisfy: angle > ±10 of 90°

• 7. Remove Weaker Edges

Remove connected edges that do not satisfy: angle > ±10 of 90°

• 7. Remove Weaker Edges

• 8. Plot Pixel Distribution

• 9. Convert to FFT Domain

Training Test

• No. of images

Edge vectors (per image) Total no. of feature vectors 90

(45HDR; 45LDR)

10

(5HDR; 5LDR)

100 100 9,000 1,000

2 classes: LDR / HDR

Essentially classifying each edge independently

• 10. SVM: Train and Classify

• Each classification from test set is mapped back to

its respective image set (100 per image)

• Majority voting of the results to yield overall

image classification

Results

• Each classification from test set is mapped back to

its respective image set (100 per image)

• Majority voting of the results to yield overall

image classification

Test Image Actual Predicted Confidence 1 2 3 4 5 6 7 8 9 10 LDR LDR 87 LDR LDR 92 LDR LDR 100 LDR LDR 91 LDR LDR 90 HDR HDR 88 HDR HDR 99 HDR HDR 80 HDR HDR 69 HDR HDR 55

Results

• A proof-of-concept has been presented for detecting HDR processed

images

• Halo artefact present in iPhone 4 HDR edges
• Large peak in intensity values characterised at strong edge points
• Strategy for detecting strong edges identified
• Scheme tested and trained on 100 images
• Classification accuracy of 100% (after majority voting)

Summary

• Strengthen strategy for detecting strong edges
• Consider edges of all possible orientations
• Greatly increase image set size
• Extend to classify from more device sources
• Classify HDR Apps that created the image

Future Work

Questions?