In this paper

a novel feature-based image watermarking scheme is proposed. Firstly

the Harris-Laplace detector is utilized to extracted feature points

which can survive a variety of attacks. Then

nonoverlapped disks around local feature points are defined adaptively according to the feature scales

Finally

several copies of the digital watermark are embedded into the nonoverlapped disks by comparing the DFT mid-frequency magnitudes. In watermark detection

the digital watermark can be extracted by using the maximum membership criterion. Experimental results show that the proposed scheme is not only invisible and robust against common signals processing such as median filtering

sharpening

noise adding

and JPEG compression

but also robust against the desynchronization attacks such as rotation

translation

scaling

row or column removal

shearing

and local random bend.