Analysis of a wavelet-based watermarking algorithm

In this paper the analysis of a wavelet-based water-marking algorithm is presented, in which Discrete Wavelet Transform (DWT) is used for watermark embedding and detection. In image compression field, the relation between DWT coefficients and Human Visual System (HVS) is researched intensively to get better compression performance. Recently this knowledge is applied to the watermarking algorithms based on DWT domain. The watermarking algorithm to analyze is adaptive to the image characteristics, such as texture and luminance contents, and it is realized using pixel by pixel masking in DWT domain to embed watermark with maximum energy. The detection process consists of cross-correlation between marked image and pseudorandom sequences, and in this process the original image is not required. The computer simulation results show robustness and perceptual invisibility of the algorithm.