Combining Mexican hat wavelet and spread spectrum for adaptive watermarking and its statistical detection using medical images

This paper present a secure medical image watermarking technique applying spread-spectrum concept in wavelet transform domain is proposed. In the first step, discrete wavelet transform(DWT) decomposes the cover medical image into four frequency sub-bands using Mexican hat as mother wavelet and then corresponding to each pixel of the binary watermark a pair of Pseudo-Noise (PN) is embedded into a horizontal (HL) and a vertical (LH) sub-band. In order to maintain the imperceptibility of the watermarked image, strength of the generated PN sequence pair is adjusted according to specified document to watermark ratio (DWR). For the extraction the watermark, statistical profile of DWT coefficients of watermarked image is determined and the obtained probability distribution function (pdf) is utilized for designing the watermark detection procedure. Proposed detector considers the best fitted Cauchy statistical model of heavy-tailed family, which accurately models the non-Gaussian DWT coefficients of an image. The robustness of the method is examined for various kinds of attacks with varying watermark to document ratio. Further, experimental results show that the proposed technique offer more robustness than other state-of-the-art method.