Vinicius Licks

Geometric attacks on image watermarking systems

Synchronization errors can lead to significant performance loss in image watermarking methods, as the geometric attacks in the Stirmark benchmark software show. The authors describe the most common types of geometric attacks and survey proposed solutions.

Angle QIM: a novel watermark embedding scheme robust against amplitude scaling distortions

Quantization index modulation (QIM) watermarking has received a great deal of attention ever since the rediscovery of Costas result on codes with host-interference rejecting properties. While such embedding schemes exhibit considerable improvement in watermark capacity over their earlier predecessors, (e.g. spread-spectrum), their fragility to even the simplest attacks soon became apparent. Among such attacks, amplitude scaling has received special attention. We introduce a quantization scheme, named angle QIM (AQIM), that is provably insensitive to amplitude scaling attacks. Instead of embedding information by quantizing the amplitude of pixel values, AQIM works by quantizing the angle formed by the host-signal vector with the origin of a hyperspherical coordinate system. Hence, AQIMs invariance to amplitude scaling can be shown by construction. Experimental results are presented for the bit error rate performance of AQIM under additive white Gaussian noise attacks.

The effect of the random jitter attack on the bit error rate performance of spatial domain image watermarking

In this paper we study the effects of synchronization errors on the bit error rate performance of spatial domain image watermarking. For that, we introduce the concept of the random jitter attack. This attack is characterized by the displacement of each pixel position by a random amount given by an arbitrary distribution, followed by interpolation over the modified sampling grid. An analogy is made between this attack and the errors introduced by timing jitter during sampling. Under such an analogy, a channel model is proposed to describe the effects of the jitter attack on the watermarked image. This model suggests that the effects of the jitter attack can be analyzed as the addition of signal dependent noise to the watermarked image. We derive expressions that relate the strength of the jitter attack to the bit error rate obtained for watermark decoding. These expressions are compared to experimental results obtained by performing the random jitter attack over a spatial domain spread spectrum watermark.

Angle QIM: on document to watermark ratio analysis

Max Costa original results on dirty paper codes have given the foundation for Quantizatin Index Modulation (QIM) based for digital image watermarking. However, the standard QIM methods can be greatly degraded by simple amplitude scaling attacks. In order to overcome this limitation, Angle Quantization Index Modulation (AQIM) schemes have been recently introduced. By quantizing the angle of the watermark vector according to a symbol dependent lattice, AQIM’s construction leads to an inherent invariance against amplitude scaling distortions. In this paper, we present a theoretical analysis of the watermark distortion introduced by AQIM during the embedding process based on the angle quantization step parameter, ¢µ. Such theoretical analysis is validated by comparison with experimental results, which are also included in this paper.

An exact expression for the bit error probability in angle QIM watermarking under simultaneous amplitude scaling and AWGN attacks

The performance of watermarking methods based on lattice quantization schemes can be greatly degraded by simple amplitude scaling attacks. Scaling the amplitude of pixel values by relatively small amounts have the potential effect of moving the watermark vector away from its original quantization centroid, thus leading the decoder to incur erroneous decisions. In order to overcome this limitation, angle quantization index modulation (AQIM) schemes have been recently introduced. By quantizing the angle of the watermark vector according to a symbol dependent lattice, AQIMs construction leads to an inherent invariance against amplitude scaling distortions. In this paper, we proceed with a thorough theoretical analysis of the two-dimensional version of AQIM, leading to an exact expression for the bit error probability for simultaneous amplitude scaling and AWGN attacks. Such theoretical expressions were validated by comparison with experimental results, which are also included in this paper.

Performance bounds on optimal watermark synchronizers

The inability of existing countermeasures to consistently cope against localized geometric attacks has precluded the widespread acceptance of image watermarking for commercial applications. The efficiency of these attacks against the so-called spread spectrum methods resides in their ability to affect the synchronization between the watermark reference and the extracted watermark at the detector end. In systems based on quantization schemes, geometric attacks have the effect of moving the watermark vector away from its actual quantization centroid, thus causing the watermark decoder to output wrong message symbols. In this paper, our goal is to gain a better understanding of the challenges imposed by the watermark synchronization problem in the context of localized geometric attacks. For that matter, we propose a model for the watermark synchronization problem based on maximum-likelihood (ML) estimation techniques. In that way, we derive theoretically optimal watermark synchronizer structures for either blind or non-blind schemes and based on the Cramer-Rao inequality we set lower bounds on the variance of these attack parameter estimates as a means to assess the accuracy of such synchronizers.