摘要

A new robust watermarking algorithm based on conjugate symmetric sequency-ordered complex Hadamard transform (CS-SCHT) is proposed in this paper. At the phase of embedding, the magnitude of the CS-SCHT coefficients of host image is selected as support of watermark, and a new human visual system is designed to improve the invisibility of marked image. Kolmogorov-Smirno method is used to estimate the approximate distribution of the coefficients' magnitude, and then a locally optimal detector is constructed based on the approximate distribution. Experimental results show that the CS-SCHT is suitable for watermark application, and the proposed detector and human visual system can provide good performance.

全文