Image watermarking in real oriented wavelet transform domain

In this paper, we propose blind and non-blind watermarking schemes in the real oriented wavelet transform (ROWT) domain. The ROWT, which is a member of the dual tree complex wavelet transform (DTCWT) family, is chosen as a watermarking domain since the DTCWT has recently emerged as an important new image processing tool. Existing watermarking schemes based on the DTCWT usually lack high embedding capacity. This is mainly due to the fact that the left inverse and the right inverse of the DTCWT (including the ROWT) are not equal. We have observed a relation when the ROWT follows its left inverse, and have used this relation to develop two watermarking schemes in the ROWT domain. Experimental results show that the proposed ROWT based watermarking schemes not only have a much higher capacity than the existing DTCWT based watermarking schemes, but are also robust to various image modification operations such as cropping, Gaussian filter, Gaussian noise, and salt and pepper noise.