In order to be more robust and with higher fidelity

watermarking should be based on capacity estimation.In this paper

two capacity estimation based watermarking algorithms are proposed.First HVS are used to decide the embedding strength

and then these strength are used for capacity estimation to decide which coefficient are used to embed information bit.The algorithms can achieve high robustness

low distortion and dynamical adjustment of embedding capacity

and can be used for binary or image watermarking.A method combining watermarking and JPEG compression is also proposed here

and shows good performance for both requirements.