Optimization of Fidelity with Adaptive Genetic Watermarking Algorithm Using Roulette-Wheel

In Digital Watermarking a low energy signal is imperceptibly embedded in another signal. The low energy signal is called watermark. Earlier techniques based on spatial and frequency domain, had the problems of a poor fidelity especially with higher payloads of watermark image. Unsatisfactory value of Peak Signal to Noise Ratio (PSNR) of watermarked image resulting in image quality degradation was seen in varying degree, in various research works. This paper is an attempt to employ genetic algorithms to find suitable locations for watermark insertion within a cover image, to focus on the issue of optimizing fidelity. The paper uses a roulette-wheel selection strategy in genetic algorithm and investigates the variation of maximum fitness which reflects the higher PSNR of the watermarked image with respect to embedding strength, number of genes, and incremental payloads of watermark consisting of binary matrix. With the proper understanding and analysis of the results of this paper, a successful watermarking scheme can be obtained using Genetic algorithms, which may be useful when optimizing the fidelity aspect of digital watermarking.