Double compression detection based on feature fusion

This paper proposes a feature fusion based method for JPEG double compression detection. When extracting the fusion features, the Markov model of the first digits of DCT coefficients and the difference of adjacent coefficients extracted from JPEG images are fused to a big feature vector, and then the Sammon mapping method is applied for dimensionality reduction. Finally, the Support Vector Machine is applied to detect the JPEG double compression based on the feature fusion. The experimental result indicates that for the images compressed by QF1=QF2 or QF1=95, our method can still achieve a relative high classification accuracy which most of the previous algorithms have reported to be powerless.

[1]  Rainer Böhme,et al.  Block convergence in repeated transform coding: JPEG-100 forensics, carbon dating, and tamper detection , 2013, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.

[2]  Florent Retraint,et al.  JPEG Quantization Step Estimation and Its Applications to Digital Image Forensics , 2017, IEEE Transactions on Information Forensics and Security.

[3]  Sam Kwong,et al.  An Effective Method for Detecting Double JPEG Compression With the Same Quantization Matrix , 2014, IEEE Transactions on Information Forensics and Security.

[4]  Wei Su,et al.  A machine learning based scheme for double JPEG compression detection , 2008, 2008 19th International Conference on Pattern Recognition.

[5]  Jan Lukás,et al.  Estimation of Primary Quantization Matrix in Double Compressed JPEG Images , 2003 .

[6]  Wei Su,et al.  A generalized Benford's law for JPEG coefficients and its applications in image forensics , 2007, Electronic Imaging.

[7]  Tomás Pevný,et al.  Detection of Double-Compression in JPEG Images for Applications in Steganography , 2008, IEEE Transactions on Information Forensics and Security.

[8]  Fouad Khelifi,et al.  On the SPN Estimation in Image Forensics: A Systematic Empirical Evaluation , 2017, IEEE Transactions on Information Forensics and Security.

[9]  Gerald Schaefer.  An uncompressed benchmark image dataset for colour imaging , 2010, 2010 IEEE International Conference on Image Processing.

[10]  Bo Wang,et al.  Double Compression Detection Based on Markov Model of the First Digits of DCT Coefficients , 2011, 2011 Sixth International Conference on Image and Graphics.

[11]  Bin Li,et al.  Detecting doubly compressed JPEG images by using Mode Based First Digit Features , 2008, 2008 IEEE 10th Workshop on Multimedia Signal Processing.