Transcript
JPEG Image Steganalysis Using Variation Technique
IIT School of Applied Technology
Stephen Danielsdaniel@hawk.iit.edu
Indhi Powletteipowlett@hawk.iit.edu
Ana Orozcoaorozco1@hawk.iit.edu
Cruz Tovarctovar@hawk.iit.edu
Steganography
Steganography is the art of passing information in a manner that the existence of the message is unknown.
Project Goals
Create JPEG images with hidden data using steganographic tools
Use Variation Technique to analyze clean and dirty images
Test variation technique validity by analyzing clean and dirty images
What is JPEG?
JPEG stores 24 bits of color per pixel = 16 million colorsJPEG is able to display rich colorsCan alter up to 3,932 pixels in a 65,535X65,535 pixel image
JPEG is a lossy compression algorithmJPEG discards some data when compressed to reduce file size
Least Significant BitIs a common technique used in Steganography
LSB is the lowest value in the bit orderLSB is replaced with the hidden data
Allows stego image to not appear significantly different from original
Variation TechniqueDeveloped by Jessica Fridrich and first detailed her book.
Statistical analysis may uncover clues of hidden data in an image.
EntropyOccurs when hidden data is
embedded into an image
Random uniform noise from the hidden data
Entropy increases when hidden data is embedded
Variation TechniqueIt is a measure of the entropy in an
image
The more hidden information in the
carrier, the amount of entropy increases
Process150 Clean Images
150 Outguess Images
150 F5 Images
Text files ranged from 65 bytes to over 9000 bytes
Steganographic ToolsOutguess
Open source programHides data in least significant bit
F5Hides data by implementing matrix encoding Improves the efficiency of embedding
ConclusionAside from Jessica, no one really has tried using the variation
technique
Analysis with variation technique shows interesting results for F5 and Outguess
Differences in result values could be attributed to different hiding methods used by F5 and Outguess