Robust Image Watermarking Using Mean Quantization in DTCWT Domain

This paper presents two quantization based image watermarking schemes in dual tree complex wavelet transform (DTCWT) domain. The energy map of the original image is first computed from the six high-frequency sub-bands of DTCWT, and the watermark is embedded into the high-energy pixels. The proposed two schemes embed the watermark by quantizing the high-frequency and low-frequency DTCWT coefficients, respectively. In our schemes, the watermark capacity can be more than 1000 bits. Experimental results show that the watermark can be extracted under JPEG compression, added Gaussian noise, filtering, image enhancement, etc.