Watermark detection is one of the most important steps in digital watermarking.It has traditionally been tackled using correlation-based techniques.However

correlation-based detection is not the optimum choice since it does not utilize the distributional characteristics of the host media.In this paper

by applying generalized Gaussian distributions to statistically model the AC DCT coefficients of the original image

a DCT-domain watermark detection scheme based on local optimum detection is proposed.The performance of the new detector in terms of the associated probabilities of detection and false alarm is analyzed.The experimental results support the theoretical performance analysis

and confirm the superiority of the novel detector at the same time.