Fireworks Algorithm and Particle Swarm Optimization for Watermarking in Image Vector Quantization

Information hiding has been addressed in several studies. When the information is hidden in digital images, it can be carried out in three domains: spatial domain (original space of image pixels), frequency domain (or transformed domain, such as Discrete Cosine Transform domain) and compressed domain. In the last case, one can cite data embedding in images compressed by vector quantization (VQ). This paper addresses the problem of codebook partition in the scenario of invisible watermark embedding in digital images compressed by VQ. In this paper, two techniques are investigated for partitioning purposes: the PSO (Particle Swarm Optimization) algorithm and the EFA (Enhanced Fireworks Algorithm) as alternatives to the Genetic Algorithm. The performance of the techniques is evaluated with regard to the algorithms execution time. Robustness of the watermark against a variety of attacks is assessed for codebooks partitioned with the aforementioned algorithms.