A robust image authentication method distinguishing JPEG compression from malicious manipulation
By Ching-Yung Lin, Shih-Fu Chang. IEEE Trans. Circuits Syst. Video Technol., 2001
607 citations recorded by the source corpus.
Subfields: Computer Vision and Pattern Recognition
View Paper: https://www.semanticscholar.org/paper/eb06303740fec013fd6dc832631fffc52c6dec85
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computer Vision and Pattern Recognition
- Min Wu
- K.J.R. Liu
- Christoph Zauner
- Ashwin Swaminathan
- Assa Naveh
- Martin F. H. Schuurmans
- Eran Tromer
- Yinian Mao
Citations in the corpus
- Digital watermarking (2002, methodology): “If X and K denote generic random variables in the random vectors XN and KN , respectively, then any dependence between X and K are modelled by the joint probability mass function, p(x, k).”
- Digital image forensics via intrinsic fingerprints (2008, background): “In doing so, some works assume that creating a tampered image involves a series of processing operations, which might include resampling [15], JPEG compression [16], [ 17 ], Gamma correction [18], and chromatic aberration [19].”
- PhotoProof: Cryptographic Image Authentication for Any Set of Permissible Transformations (2016, background): “…header [30], or in digital artifacts (see [16] for a survey), but in general this can be time-consuming and unreliable.2 An alternative approach, pursued in this paper, is to associate additional data (a signature or proof) with the final image, in order to detect forgery reliably and robustly.”
- Implementation and Benchmarking of Perceptual Image Hash Functions (2010, background): “[24] shows that a feature code can be extracted from the relationship between two DCT coefficients of the same position in two separate blocks.”
- Robust and secure image hashing (2006, methodology): “Some of the features that have been proposed in the literature include block-based histograms [24]–[26], image-edge information [27], relative magnitudes of the discrete cosine transform (DCT) coefficients [28], and the scale interaction model with the Mexican-Hat wavelets [29].”