Double Compression Detection Based on Markov Model of the First Digits of DCT Coefficients

Dong, Lisha; Kong, Xiangwei; Wang, Bo; You, Xinge

doi:10.1109/icig.2011.100

Cited by 10 publications

(15 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To localize tampered regions in the presence of double compression, Li et al proposed a sliding-window image forgery detector [21] for improving the performance of the saliency map-based tampering detector of [22]. Contrary to the approaches like [20,27,33] which are exploited the first order statistics, by using a Markov transition probability matrix, two methods [5,6] utilized the second-order statistics of QDCT coefficients to enhance the performance of double compression detection algorithms by leveraging the inter-block correlation of coefficients.…”

Section: Related Workmentioning

confidence: 97%

“…In order to predict the condition whether a query JPEG image is double compressed or not, different methods have been developed [2,5,6,14,20,23,26,27,31,33]. The ideas behind double compression detection techniques can potentially facilitate the localization of altered regions in images [3,9,21,22,38].…”

Section: Related Workmentioning

confidence: 99%

“…As the second category, due to the nature of non-overlapping block-based processing in JPEG encoder, there are double compression detection methods that leverage inconsistencies introduced in the boundaries of image's blocks or the lack of correlation between neighboring blocks [5,6,23]. In the third category, the statistical distributions of quantized discrete cosine transform (DCT) coefficients in the compressed domain are investigated [5,6,20,27,27,33].…”

Section: Related Workmentioning

confidence: 99%

“…Provided that the utilized methodology in JPEG compression scheme is explicitly known, the secondary quality level, l 2 , is also uniquely identifiable. In such a scenario for which analyzer is aware of the secondary quality level, but unaware of the primary quality level, l 1 , several semi-aware techniques have been developed to distinguish single compressed images from double compressed ones [5,6,20,26].…”

Section: Motivating Forensic Scenariomentioning

confidence: 99%

“…Because of proper matching between Benford [1] and 2-D DCT coefficients distributions [30,36], some double compression detection techniques have been analyzed the cue of quantized DCT (QDCT) coefficients statistics of doubly compressed JPEG images [6,20,27]. In [20], Li et al proposed a feature extraction approach based on the first digit attributes of individual alternating current (AC) modes from quantized 2-D DCT coefficients in a zig-zag order.…”

Section: Related Workmentioning

confidence: 99%

See 4 more Smart Citations

Quantization-Unaware Double JPEG Compression Detection

et al. 2015

View full text Add to dashboard Cite

The current double JPEG compression detection techniques identify whether or not an JPEG image file has undergone the compression twice, by knowing its embedded quantization table. This paper addresses another forensic scenario in which the quantization table of a JPEG file is not explicitly or reliably known, which may compel the forensic analyst to blindly reveal the recompression clues. To do this, we first statistically analyze the theory behind quantized alternating current (AC) modes in JPEG compression and show that the number of quantized AC modes required to detect double compression is a function of both the image's block texture and the compression's quality level in a fresh formulation. Consequently, a new double compression detection algorithm is proposed that exploits footprints introduced by all non-zero and zero AC modes based on Benford's law in a low-dimensional representation via PCA. Then, some evaluation frameworks are constructed to assess the robustness and generalization of the proposed method on various textured images belonging to three standard databases as well as different compression quality level settings. The average F 1 -measure score on all tested databases in the proposed method is about 74 % much better than the state-of-theart performance of 67.7 %. The proposed algorithm is also applicable to detect double compression from a JPEG file and localize tampered regions in actual image forgery scenarios. An implementation of our algorithms and used databases are available upon request to fellow researchers.

show abstract