Frequency Domain for Color Image Authentication Proofing

Abstract

Digital watermarking is the process of embedding data known as a watermark into a multimedia object so that the watermark can later be identified or recovered to claim its object. The use of an embedded watermark would enable the owner of the work hardware to be identified. In the proposed method the Particle-Swarm-Optimization (PSO) algorithm has been applied to embed a binary watermark bit in the optimal color sub-bands, resulting in a less extreme value after the PSO point. The embedding level, which uses Discrete Wavelet Transform (DWT) to embed binary bits in the resulted image in the low-frequency domain, must be able to withstand intended attacks while maintaining high quality, as well as assess image authentication and localize tampering areas (s). Various consistency metrics and watermark removal attacks are used to evaluate the algorithm in an experimental environment. The results show that the model can support imperceptible watermarking as well as high attack resistance.