An Optimized Alpha-blending watermarking technique in DWT domain for digital color images

Watermarking is the process of‎ hiding digital information in a carrier signal like image‎, ‎audio‎, ‎video‎, ‎or any‎ other media system in its work environment‎. Due to copyright protection, recently, many researchers try to propose new efficient watermarking methods. ‎In this paper‎, ‎a robust image ‎watermarking technique has been implemented for the copyright protection‎ based on 3-level discrete wavelet transform‎. ‎In this approach‎, ‎a multi-bit watermark is embedded into the low-frequency sub-band of a cover image by using‎ Alpha-blending technique with optimized coefficients‎. Accordingly‎, ‎a criterion‎ for the distance between two matrices is defined initially‎. ‎Then according to the‎ defined criteria‎, ‎optimal values for parameters of the Alpha-blending method‎ are presented using the concepts of mathematical optimization and the GAMS‎ software‎. ‎The efficiency of the proposed method is evaluated by using statistical parameters such as peak-signal-to-noise ratio (PSNR) and mean-square‎ error (MSE)‎. ‎The experimental results show that while the watermark images‎ generated by the proposed algorithm are invisible‎, ‎the quality of watermarked‎ image‎, ‎retrieved image and the recovered image have been also improved‎.