Fragile and Blind Reversible Watermarking Method for Color Images using 4-Level DWT

摘要

Digital watermarking algorithms are widely applied to digital color image for ownership protection and tampering detection. Digital watermarking is the process of reversibly or irreversibly embedding information into a digital signal. Blind reversible watermarking has an additional advantage of recovering the image which is same as the original image pixel by pixel, after the image is authenticated. A fragile watermark is distorted in the case of tampering, notifying an investigator upon extraction of potential malicious manipulation in the image. A new algorithm is suggested for reversible watermarking which makes use of difference expansion concept in this paper. In the suggested algorithm, 4 level DWT of the input color image is calculated and the watermark bits are embedded in the transformed coefficients which don’t cause any overflow or underflow. The result for embedding up to 8 watermark bits per coefficient is presented. The complete VLSI implementation of watermark embedding and extraction is presented on the FPGA SPARTAN 3E board.