Improved blind DCT watermarking by using chaotic sequences

This paper proposes a modified blind DCT watermarking method for grayscale images, by the use of chaotic maps. The proposed method is resilient to certain watermarking attacks (compression, filtering, noise addition, cropping etc.) and represents an improvement in terms of security for Zhao et al.'s method. Furthermore, we propose a watermark removal attack, which succeeds against the standard method of Zhao et al., but has low reliability against the proposed chaotic attack.