Adaptive color image watermarking algorithm based on two-dimensional chaos in space and time

digital watermarks;  space-time chaos;  human visual models;  Discrete recombination transformation domains;  Robustness

An adaptive color image watermarking algorithm based on space-time chaos in the discrete cosine transformation domain (DCT domain) is proposed. The algorithm uses the spatiotemporal chaos model to generate chaotic sequences to scramble and encrypt the binary watermark, and determines the watermark embedding intensity by combining the improved human vision model (HVS) image block classification method and the RGB component characteristics of color images, and finally the watermark is repeatedly embedded in the DCT intermediate frequency transformation coefficient of each component of the color image RGB. Experimental results show that the algorithm has very good robustness against various noise, filtering and compression attacks, and watermarks with similarity of 30.25 and 0.914 can still be extracted when subjected to attacks with a compression ratio of 0% and a shear of 869%, respectively.