ABSTRACT
This chapter discusses a variety of bio-inspired algorithms that are employed by digital image watermarking for optimum selection of parameters. The performance of the optimization algorithms is verified by collecting the mathematical convergence proofs and by testing over the benchmark problems. The watermark embedding into a host image and extraction from a test image are done iteratively using conventional watermarking algorithms and Genetic programming/genetic algorithm generated scaling factors/locations/intensities. Swarm algorithms are the most recent and widely used bio-inspired algorithms for finding optimized solutions. The tabu search algorithm is based on the process designed to cross boundaries of feasibility or local optimality instead of treating them as barriers. Various image watermarking techniques based on firefly algorithm are proposed in combination with image processing transforms such as discrete cosine transform, discrete wavelet transform, singular value decomposition, and QR decomposition. The fitness functions are calculated using different evaluation parameters of watermarking which are related to quality and robustness.
