Security of copy-move forgery detection techniques

Nguyễn, Hiếu Cường; Katzenbeisser, Stefan

doi:10.1109/icassp.2011.5946869

Cited by 8 publications

(4 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(a) Copy-moved (b) Block-based detection [6] (c) SIFT-based detection [9] (d) geometric attack [12] on (a) (e) Block-based detection (f) SIFT-based detection The interpretation of such results is straightforward: SIFT features have been devised in such a way to be robust against geometric manipulations, which obviously fail at impairing a SIFT-based copy-move detector. On the other hand, the classification-based attack is designed to preserve the local visual quality of the image.…”

Section: Block-based and Sift-based Copy-move Detectionmentioning

confidence: 99%

“…We selected the forged image of Figure 14a, where a person has been removed from the picture by duplicating a portion of the sandy region. Starting from it, we hid the copy-move forgery by means of the following counter-forensic techniques: the classificationbased attack and the geometric attack proposed in [12]. According to the authors, the attack of [12] proved to be effective against a number of block-based detectors (including [6]).…”

Section: Block-based and Sift-based Copy-move Detectionmentioning

confidence: 99%

“…Starting from it, we hid the copy-move forgery by means of the following counter-forensic techniques: the classificationbased attack and the geometric attack proposed in [12]. According to the authors, the attack of [12] proved to be effective against a number of block-based detectors (including [6]). It consists of a crop of 3 pixels (columnwise and row-wise), followed by two JPEG compressions (qualities 70 and 60), and a final resampling (bicubic interpolation) back to the size of the original image.…”

Section: Block-based and Sift-based Copy-move Detectionmentioning

confidence: 99%

“…Chrislein et al [11] studied the robustness of several detectors against common image processing and observed that block-based methods are not robust against geometric manipulations (e.g., resampling, cropping), lossy compression, and noise addition. Nguyen et al [12] successfully impaired three well-known block-based detectors by combining some of these manipulations into a simple yet effective counterforensic scheme (see [12]). The same results, however, cannot be replicated with SIFT-based detectors inheriting the intrinsic robustness against geometric transformations from Lowe's algorithm.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Counter-forensics of SIFT-based copy-move detection by means of keypoint classification

et al. 2013

View full text Add to dashboard Cite

Copy-move forgeries are very common image manipulations that are often carried out with malicious intents. Among the techniques devised by the 'Image Forensic' community, those relying on scale invariant feature transform (SIFT) features are the most effective ones. In this paper, we approach the copy-move scenario from the perspective of an attacker whose goal is to remove such features. The attacks conceived so far against SIFT-based forensic techniques implicitly assume that all SIFT keypoints have similar properties. On the contrary, we base our attacking strategy on the observation that it is possible to classify them in different typologies. Also, one may devise attacks tailored to each specific SIFT class, thus improving the performance in terms of removal rate and visual quality. To validate our ideas, we propose to use a SIFT classification scheme based on the gray scale histogram of the neighborhood of SIFT keypoints. Once the classification is performed, we then attack the different classes by means of class-specific methods. Our experiments lead to three interesting results: (1) there is a significant advantage in using SIFT classification, (2) the classification-based attack is robust against different SIFT implementations, and (3) we are able to impair a state-of-the-art SIFT-based copy-move detector in realistic cases.

show abstract

Section: Block-based and Sift-based Copy-move Detectionmentioning

confidence: 99%

Section: Block-based and Sift-based Copy-move Detectionmentioning

confidence: 99%