Adaptive Wavelet-based Family Tree Quantization for Digital Image Watermarking

This paper presents an adaptive wavelet-based blind digital watermarking scheme for copyright protection. The wavelet coefficients in the same spatial directions at different decomposition levels are grouped into family trees. The watermark is embedded by quantizing the family trees. The trees are adaptively quantized using the characteristics of the human visual system in the wavelet domain so the maximum allowable changes can be made to the original image without any visible distortions. In the meantime, the quantized trees exhibit a large enough difference for blind watermark extraction. Experimental results demonstrate that the proposed system achieves comparable performance as the fixed quantization approach. Specifically, the proposed system is robust against common image processing attacks and more resistant to the compression attacks than the fixed quantization approach.