A robust watermarking scheme exploiting balanced neural tree for rightful ownership protection

作者：Rani Asha; Raman Balasubramanian; Kumar Sanjeev^*

来源：Multimedia Tools and Applications, 2014, 72(3): 2225-2248.

DOI：10.1007/s11042-013-1528-3

• Discrete wavelet transform
• Neural tree
• Semi-blind watermarking
• Synapses
• Watermarking attacks

This paper presents a novel semi-blind watermarking scheme in which a balanced neural tree (BNT) is exploited for embedding and extracting the watermark. The BNT is trained to learn the watermark and the synapses (optimal weights) of the trained BNT are embedded in the host image instead of watermark image. As a result, the proposed scheme is able to embed a large size of watermark in the host image. In the watermark extraction phase, the embedded synapses are extracted from the transmitted image and then watermark is recovered from these extracted synapses. In this way, the original image is not required in the extraction phase. The proposed scheme is tested to withstand various kinds of attacks and found to be robust against various image processing and geometrical attacks.

• 出版日期2014-10